Quality is often understood as “strength for application”, but such a concept has no cost or processing considerations. A part produced using poor processes, which can be guaranteed only as the result of inspection, has no inherent quality. A fuller definition is “dependability of function at lowermost cost, consequential from good design and the use of skilled processes”.

Quality begins on the understanding application and on creating the drawing. With awareness of rubber properties and its manufacturing needs for simple design to optimize functions and avoid manufacturing difficulties towards tooling and finished product processing. Such components will have best values for money.

Using the knowledge and understanding of the rubber designer at an initial stage will enhance the design which will offer rubber’s optimum properties and avoid future problems. It will also allow considering the tolerances, differences of thermal expansion, co efficient of shrinkage etc…

The goal of all quality conscious companies is zero defects. This is achieved through the use of capable processes and statistically based monitoring; it cannot be achieved by inspection.

Continual improvement is needed to increase the capability of processes. In today’s rubber industry, this capability is relatively low and many aspects are difficult to monitor. For example, rubber is pliable and dimensions such as cylindrical diameters often cannot be gauged quickly and accurately. In mixing, dispersion cannot be easily checked; in preparing blanks, actual volume cannot easily be measured; and in moulding, the injection process is often more sensitive to material variation than any existing rheometer.

Rubber materials crosslink and are time and temperature sensitive. This results in significant viscosity variations during flow. Friction heat through mould gates and channels creates differences between material and mould temperatures which cannot easily be measured.

Despite such problems, much progress has been made in recent years through the emphasis on consistent processing and with modern computer-based injection presses which are self-adjusting, within limits, to material variations.

At present, world class levels of internal rejects should not be assumed in rubber but they can be achieved through careful design and adherence to the procedures outlined above. Across the industry, internal reject rates run at a typical but unacceptable level of 2%, but rates of 500 ppm (0.05%) – well within world class levels (Anderson Report) – can be sustained across a range of parts, materials and machines.

Clearly, any single part with well designed processes can achieve lower ppm reject levels. As more and more parts are designed on the right basis, and continuous improvement is applied to processes, capabilities will rise and reject levels fall.

It is important to consider present capabilities. The following is intended to provide a general guide using typical results, but every material, machine and process has its own capability and these can vary considerably.

Hardness is measured by pressing an indentor into the rubber and measuring penetration. Shore A is based on an immediate reading using a spring applied load. Variation from user to user can be as great as +/-2.5°. IRHD uses a dead load with a 30 second wait and is more consistent, giving user variations of +/-1.5° (3 sigma). There can be significant differences between the two types of readings.

The Tolerances Table (BS 3734) gives a brief background to rubber moulding tolerances. Tighter limits can be achieved (particularly with injection moulding) by slowing the process and moulding under minimum stress conditions. Limiting the number of cavities and shortening the flow path for the rubber also gives improvements, but clearly all these measures have a commercial cost.

Tolerances in rubber are generally less critical as the material deforms readily and accommodates variations. In fact, errors in measuring rubber can be significant and non-contact methods should be used wherever possible.

When designing tooling, critical dimensions should be taken into account to minimise the effects of tool split lines and flow. Stresses built up while the material flows can be moulded in if curing begins before the rubber is relaxed. On removal from the mould the part will distort accordingly, resulting in lower dimensional capabilities.

The Tolerance Table for BS 3734 moulding tolerances are included in this site. M2 tolerances are normal commercial tolerances and can be met by most rubber materials. M1 tolerances can be achieved through careful design and consideration. Production rates may be affected in some instances and good tooling and equipment is required. Particular attention should be given when using high shrinkage materials such as Silicones, Fluoroelastomers and peroxide cured rubbers.

Keypad force tolerances should not be considered in the same way as dimensional tolerances. Nearly all keypads are finger operated and the finger is relatively insensitive to exact loads. This is particularly true of single finger operation where loads of +/- 30% will not be noticed by a user. Examples include car switches, teletext terminals, car phones and instrument controls.

The exceptions are computer and typewriter keyboards where a tolerance of +/- 15% is required. In this case the finger rarely reaches full travel and operators are sensitive to forces because of the high frequency of use.

Capabilities depend upon membrane design and length of travel. Typical capabilities are outlined in this table.

Force variation is greatly affected by the fixing of the keypad base and by the escutcheon/keycap interface with the rubber. It is strongly recommended that early advice is sought.