Difficult Moulding – Tight Tolerances

Metflex has applied SPC to achieve high moulding precision in rubber manufacturing.

Rubber by its very nature can be difficult to mould to very tight tolerances. Leaving aside environmental conditions, the shape of the pre-moulded blank, the viscosity of the rubber, the condition of the tool and press, moulding tool configuration, and even a change of press operator, can all have a bearing on the tolerances of the finished moulded product.

Understanding how these factors affect the dimensional accuracy of precision moulded parts has prompted Metflex Ltd to introduce statistical process control (SPC) monitoring techniques to consistently achieve unprecedented levels of moulding precision.

Moulding Tolerances

Two years ago, Metflex observed an unexplainable variation in the dimensional accuracy of its gas meter diaphragm mouldings – the thickness of the diaphragm had increased. Domestic gas meters operate by the action of a diaphragm constantly flexing, alternately filling and expelling gas, to produce a constant gas flow. The thickness of the diaphragm is thus critical to the meter’s performance and diaphragm service life, typically 20 -30 years.

Although still within permitted tolerances, the inability to explain the variation in thickness highlighted the fact that traditional process control and inspection techniques were unable to provide the fine grain analysis needed.
The decision to introduce Statistical Process Control was based on the premise that ‘There can be no action without analysis, and no analysis without measurement’.

Statistical Process Control

Statistical Process Control (SPC) allows Metflex to collate data from a variety of production and metrology sources to monitor the moulding thickness. It uses statistical methods to monitor and control a process, with the aim of achieving optimum process conditions. Its main advantage over ‘inspection’ is that it allows early detection of a problem, thus making it easier to identify the cause.

Metflex has implemented WinSPC statistical process control software. It produces Control Charts that graphically represent the variability in a process over time, making it easier to spot inconsistencies and unnatural fluctuations.

Control charts typically display the limits that statistical variability can explain as normal. If the process is performing within these limits, it is said to be in control; if not, it is out of control. The area between each control limit and the centerline is divided into thirds. The third closest to the centerline is referred to as Zone A, the next third is Zone B, and the third closest to the control limits is Zone C. Note that there are two of each of the Zones, one upper and one lower.

Where 3-sigma is used for the control limits, Zone A is called the “1-sigma zone”, Zone B is “2-sigma zone”, and Zone C is “3-sigma zone”.

The 3-sigma rule states that nearly all values lie within three standard deviations of the mean in a normal distribution. As a result 68.3% of results (1-sigma / Zone A) are within one standard deviation, 95.5% of results (2-sigma / Zone B) are within two standard deviations, and nearly all values are within 99.7% of the standard deviation (3-sigma / Zone C).

Achieving Consistent Dimensional Accuracy

Working with its customer, Metflex established the critical dimensional parameters for the gas meter’s diaphragm and seals. The diaphragm is measured at 16 points, the seal at 8 points. At the thinnest point, the rubber moulded part is less than 2mm thick. Moulded parts are collected over an 8-hour production cycle. Four mouldings are measured per hour. The data collection is semi-automated such that thickness measurements are automatically entered into the SPC software.

The SPC control chart shows how moulding thickness varies over time. The centre line is the ideal. Within Zone A (Green) the parts are being moulded to a high degree of dimensional accuracy (nominal spec). Moving into Zone B (Yellow) indicates the parts are within tolerance, but the process needs to be tweaked to move the tolerance back into the green zone. If the results go into the Zone C (Red) – production is stopped and action taken to investigate the problem. Sharp movements/variation in thickness pinpoint where the process has changed.

Insights into the Moulding Process

Statistical process control has allowed Metflex to gain insights into the rubber moulding process that are simply beyond the reach of traditional process / inspection techniques.

Control chart A shows how the slightest change in the processing conditions can be quickly detected using SPC.

L to R – Initially there is a large variation in the thickness of the moulded part within the 1-sigma (Green) zone, eventually falling into the 2-sigma (Yellow) zone. By adjusting the moulding pressure – a known variable that can affect part thickness –

Statistical Process Control Chart A

Control Chart A

the thickness is brought closer to the centre line. Increasing the moulding pressure to 2,100psi retrieved the situation. Reducing it to 1,700psi proved to the optimum moulding pressure for this particular press, enabling a more consistent moulding.

Production Plant Efficiency

Moulding presses are used to manufacture a variety of parts. Over time the press will need maintaining as part of a preventative maintenance schedule. For the rubber moulder wishing to maximise manufacturing productivity, this may not be the best guide to press maintenance. The introduction of SPC has enabled Metflex to take an informed approach to press maintenance. Control Chart B shows the effect of press refurbishment on the thickness of the moulded part.

L to R -. Here the press initially performed well, but gradually moved into the 2-sigma zone (Yellow),

Statistical Process Control Chart B

Control Chart B

providing early warning of moulding problems. With all other processing conditions optimised, it soon became clear that the problem lay with the condition of the press. By refurbishing the press, the thickness of the moulded part was reduced from the top end of the tolerance limit to a point closer to the centre line (nominal spec).

Pushing the limits of rubber moulding

As the two Control Charts show, the application of SPC techniques has allowed Metflex to fine tune the moulding parameters to achieve greater consistency in the dimensional accuracy of its gas meter diaphragm and other rubber mouldings.
The company is now taking this a stage further to begin manufacturing even thinner rubber parts at tolerance levels far in excess of accepted industry norms such as BS 3734 : Part 1 : 1997 for rubber part manufacture. Application of SPC techniques will ensure Metflex is able to consistently meet the higher tolerance requirements needed for these ultra-thin parts.

This entry was posted in Uncategorized. Bookmark the permalink.