Simple and Efficient – Quality Assurance for Powder Injection Moulding
Download original article as a PDF
Preventative quality assurance through the testing of feedstock prior to part production can offer significant cost savings thanks to a lower reject rate, suggests Arburg GmbH + Co KG, Lossburg, Germany. Hartmut Walcher explains how the company’s SELOGICA control system can be used to provide essential quality data for PIM part producers.
In order to quickly and reliably assess feedstock batches prior to the volume production of PIM parts, the pressure at the switch over point of injection moulding machines provides meaningful, dependable values. These values enable the reliable assessment of possible fluctuations in batch quality. This data can be collected with ease using the Selogica control system of Allrounder machines.
In principle, powder injection moulding (PIM), used for the production of metal (MIM) or ceramic (CIM) components, does not differ significantly from the injection moulding of plastic parts. With these applications, the greatest possible consistency and highest quality of production must be ensured, based on predefined setting parameters.
The main challenge here is that an assessment of part quality is only possible following debinding and sintering. Quality problems in the production process can no longer be rectified after completion of these processing steps and a preliminary assessment of feedstock quality is therefore of decisive importance.
Costly assessment of moulded part quality
A period of several days is always required for the time-consuming subsequent processing of PIM parts, which poses the following problem for feedstock processors: Should they trust that the batch is satisfactory and continue to produce green compacts? Or should they stop the machines until the results of the quality testing become available?
Both options are problematic from an economical standpoint. Of course, production must run as smoothly as possible from the outset. Moreover, feedstocks, be they metal or ceramic compounds, are comparatively expensive. Copper, titanium and precious metals, for example, are processed as well as steel alloys in MIM and three digit material prices per kilo are no exception.
Added to these costs are the machine-hour rate and the hourly rates for debinding and sintering. Accordingly, the production of reject parts is always extremely problematic because of the significant costs incurred.
Many factors influence part quality
In order to ensure the flawless volume production of PIM parts, the injection moulding machine must operate perfectly and the dimensional tolerances of the cylinder, screw and non-return valve must be adhered to.
Furthermore, constant ambient conditions must be maintained to ensure production at balanced temperatures and unvarying viscosity of the feedstock, which is a central factor during processing. The viscosity depends on numerous factors, including the binder, the binder-agent content and particle size distribution, as well as the precision of feedstock preparation.
Various testing methods in use
Because of the larger number of influencing factors, it is critically important that batch fluctuations be detected prior to the actual start of production so that processors can respond quickly by implementing suitable measures or rejecting a particular batch.
In the case of feedstocks, incoming goods inspections are neither as easy to perform nor as widespread as with thermoplastics. Setting up a laboratory with a high-pressure capillary viscosimeter represents a high five-digit figure investment. This option is only viable to a limited extent as the results cannot be measured under the actual pressure and speed conditions prevailing in the injection moulding machine. In comparison, determination of the mould flow index would make more sense because of the more modest investment required, however this option can be ruled out due to its poor informative value.
An interesting alternative is the production of test pieces on a laboratory scale. For this purpose, a small standard hydraulic injection moulding machine such as the Allrounder 170 S machine could be used (Fig. 1). On an Allrounder operating under balanced temperature conditions, the maximum injection pressure at the switch over point to holding pressure can be precisely determined via the Selogica machine control system.
The test mould used produces test pieces, whereby the cavity is filled at various injection speeds and to less than 100% in each instance. The control system records the pressure at the switch over point in relation to the respective speeds as a measured value. This allows information on the quality of the feedstock to be derived from the injection moulding machine pressure values during injection. This information can then be applied to the setting parameters. Here, it is the flow characteristics of the material which are tested.
A solution that saves time and costs
The relevant tests were performed on an Allrounder 170 S. During testing, the test pieces simply have to be injected as previously described. Debinding and sintering of the parts is dispensed with completely and the green compacts can immediately be regranulated and the material reused during subsequent volume production.
Significant information on feedstock quality can be derived from the pressure recordings in relation to the respective injection speeds. Both ceramic and metal feedstocks with different binders and varying preparation requirements can be tested (Figs. 2-4).
The recording of 20 to 30 cycles is generally sufficient in order to achieve meaningful reference curves. The purchase of a small injection moulding machine with the Selogica control system is far more cost-effective than that of a high-pressure capillary viscosimeter. Moreover, a viscosimeter measures the viscosity of the feedstock at different pressures and speeds than those prevailing in the injection moulding machine.
In other words, the injection moulding machine should only be used here as a measuring instrument in a quality assurance laboratory. Although the recording of the pressure curves described here is possible using any Allrounder from production, the necessary cleaning and conversion of these machines is considerably more costly and running production is interrupted.
Test procedure and results
The correlation between feedstock quality, i.e. viscosity and the pressure/speed curves of the control system is extremely sensitive. Changes in material viscosity have an immediate impact on the characteristics of these curves.
The result is simple and clear: The switch over point is a significant differentiation criterion for the viscosity of feedstocks and consequently for their processing quality. A quick feedstock test is essentially possible on any mould, particularly as no internal pressure measurement is required for this purpose.
The most effective method, however, is the use of a small injection moulding machine operating independently of the production process. The reject rate can be effectively reduced and manufacturing costs lowered significantly. The differences between different material batches can be determined easily and reliably via the new, control-based Arburg methodology.
Calculations reveal savings potential
Based on a moulded part weight of eight grams and the use of a 316L feedstock with an average kilo price of €22, the following cost situation results for a MIM batch oven. If a four cavity mould and a cycle time of 30 seconds are used, 8.25 production hours are required to fill a hypothetical sintering oven. The feedstock costs for a throughput time of 24 hours amount to around €696 for one sintering batch.
Added to this are operating costs of around €1,200 for the debinding and sintering stations and the operating costs for the injection moulding machine, amounting to €123.75. If the quality of the feedstock cannot be tested in advance and the complete batch is not produced flawlessly as a result, total costs of €2,019.75 are incurred, which have to be written off as reject part production.
Comparable results are obtained in the case of a CIM batch oven filled with parts made from a ZrO2 (zirconium oxide) feedstock (kilo price: €120) with the same part weight. Based on a four-cavity mould and a cycle time of 30 seconds, the feedstock costs for one sintering batch amount to some €1,420, the operating costs for the debinding and sintering stations are €900 and those for the injection moulding machine are €46.25. The total costs for a normal run therefore amount to €2,366.25, which are incurred even if the batch produced is defective.
These figures impressively demonstrate that the purchase of an injection moulding machine for the purpose of preliminary feedstock testing can be recovered within a very short time if volume production thereby consistently achieves the required quality levels.
The development department for powder injection moulding at Arburg has devised a machine-based testing method for feedstocks which works quickly and simply. A batch test prior to the start of production clearly indicates whether parameter changes have to be performed for a material batch, enabling the production of PIM parts to proceed smoothly.
If the data is collected and compared over a prolonged period, acceptable batch-fluctuation ranges can be determined without the need for injection moulding parameter changes. New injection parameters can then be devised for batches which lie outside this range. This preliminary testing not only brings considerable potential for time-saving, but also successfully lowers costs thanks to the effective determination of the quality of the materials used.
Volume production progresses smoothly and the production of expensive rejects becomes a thing of the past. All the Selogica machine control systems already feature the software required for performing feedstock tests of this kind as standard. PIM processors can therefore make comprehensive use of this control option at a very reasonable cost.
ARBURG GmbH + Co KG
Postfach 1109, D-72286 Lossburg
Tel: +49 (0)7446 33 0 Fax: +49 (0)7446 33 3365
Email: [email protected]