Sponsored By

July 20, 1999

5 Min Read
The Representative Cavity:Quality Assurance for Multicavity Moulds

The process optimization aids developed for single-cavity moulds are too complicated to apply to each cavity in most multicavity moulds. Similarly, quality assurance tools that are used for single-cavity moulds, such as SPC or CPC, have only limited applicability to multicavity moulds, at least from the viewpoint of 100 percent control. Together, a mouldmaking consultant and an R&D team at the University of Paderborn, Germany, have looked into the possibility of selecting one cavity as "representative" of the mould to avoid the problem of 100 percent control, and to offer a simpler approach to quality assurance.

The idea behind the principle of the representative cavity is that cavities of a multicavity mould will differ. Some of the reasons for those differences are:

  • errors in rheological, thermal, and mechanical design,

    manufacturing tolerances during mouldmaking,uneven mould cooling, andventing problems.

These variations result in differences in quality of the parts produced, as well as in the quality of the cavities. Cavity quality may be influenced by a variation in operating point or by process disturbances. If the relative quality of the cavities is operating-point independent, it is possible to optimize the operating point and monitor the process with reference to the "worst" cavity, thus reducing the test effort by 1/n for an n-cavity mould. The worst cavity, like a "worst-case scenario," thus becomes a quality-determining, or representative, cavity for the entire mould and represents the quality level.

Finding the R-cavity

Once the representative cavity has been determined experimentally, the operating-point optimization and process monitoring can be carried out on this cavity as representative of the entire mould. The representative cavity, r-cavity for short, can be determined in various ways.

  • Complete injection shots can be made, the quality of all mouldings tested, and the r-cavity detected. The number of samples required depends on the process characteristics and the disturbances occurring. In addition, all the quality features of all moulds of a shot have to be tested.

    Alternatively, a statistical filling study plan can be used to determine the cavity with the worst filling characteristics. It is assumed that the cavity of a multicavity mould that is the last to be filled has the greatest sensitivity to disturbances and changes in process parameters. This process-sensitive behaviour is manifested primarily as shrinkage and warpage, sink marks, and voids in the moulding.Figure 3 shows that the different cavities may have different effective and absolute mould filling times.

Determination by Fill Study

Polypropylene wheel rims (Figure 1) are clad with thermoplastic elastomer in a second injection moulding process to form runners. The rims are produced in a six-cavity cold runner mould (Figure 2). The r-cavity was determined by a conventional filling study. Because the injection volume increases gradually, the mould filling process can take place after the injection of the moulding compound into the mould.

Figure 4 shows the proportions of the analyzed parts of the individual cavities as percentages of the total shot weight (excluding the sprue) plotted against the number of the filling study. The curve shape in the upper part of the figure indicates the degree of mould filling of the total shot at the respective points in the experiment. The degree of filling is calculated from the ratio of the actual shot weight to the maximum achievable shot.

Figure 4 makes clear that, in each phase of mould filling, cavity 1 contributes least to the overall shot weight. At low degrees of filling there are very significant differences between the cavities. These differences are reduced with the onset of volumetric mould filling, but are still identifiable. An "ideal" mould with identical cavities would have the proportions by weight in all the cavities at any time during mould filling (broken line).

Figure 5 shows the relationship between the differences between onset of filling and filling time for other process parameters in the current practical example. Here, the injection moulding cycle is plotted in the form of a bar chart. Starting with the machine cycle (lowermost bar), which is determined by the operating-point setting by the machine setter, the following dependencies can be identified:

  1. Effective injection times are different as a result of the various dead times at the onset of cavity filling.

    The effective times required for volumetric filling of the individual cavities, which were determined by a filling study, are achieved only for cavities 2 and 5. All other cavities are volumetrically filled under holding pressure.The effective holding pressure times are different as a result of 2.The dead times mentioned in 1 lead to different sealing times of the individual cavities.

These dependencies show that there may be significant differences in the quality features, analogous to the dependencies known from single-cavity moulds. On the other hand, it is also clear that the last cavity to be filled is "decoupled" from the rest of the mould, and thus forms a starting point for optimization and monitoring the operating point. Although operating-point optimization of this kind cannot eliminate defects introduced during the design and making of the mould, it should be considered how much it costs to modify the mould. By comparing the total quality value feasible with the existing mould with the total quality value that is theoretically obtainable (with an optimized mould), processors can decide whether the operating-point optimization provides an acceptable overall quality level, or whether the mould will have to be re-machined or rebuilt.

Operating-point Optimization

In the present example, by optimizing the operating point by means of the r-cavity according to the method described, it was possible to improve quality for the critical quality characteristics of the wheel rim ? namely hub width, hub diameter, and degree of overfilling ? from 56 to 81 percent of the overall quality theoretically achievable (all relevant single quality characteristics of a part reach the required value).

To provide a basis for assessing how representative this result is, optimization by means of the r-cavity was followed by an optimization taking into account all the cavities in the mould. This yielded the same results as in the afore-mentioned procedure, i.e., the same combination of setting parameters was calculated as in the above case.

The difference between the overall quality value (81 percent) achieved and the quality value theoretically possible (100 percent) can only be eliminated by making corrections to the mould.

Contact Information
University of Paderborn
Kunststofftechnologie (KTP)
Mr. Theodor Wenniges
Pohlweg 47-49
D-33098 Paderborn
Germany
Phone: (49) 5251 60-2227
Fax: (49) 5251 60-3821

Sign up for the PlasticsToday NewsFeed newsletter.

You May Also Like