CTI and dielectric strength of moulding materials

Two failure modes, two tests

An insulating plastic can fail in two ways. A conductive path can form across its surface, usually along a track of contamination and moisture, and the current then flows where the design assumed it could not. Or the material can break down through its thickness under a voltage high enough to exceed its dielectric strength.

The first failure is a tracking failure, described by the comparative tracking index. The second is a breakdown failure, described by dielectric strength. They are measured with different tests, they depend on different variables, and a material that is good at one is not automatically good at the other.

Confusing the two leads to the wrong material call. A housing can have an excellent dielectric strength and a modest tracking index, which matters if the design puts two live parts close together on a contaminated surface.

A design review should ask which of the two failures the product is exposed to, then check the drawing for the input that controls it. Most spacing arguments come from a design that answered one question and assumed the other.

What CTI measures

The comparative tracking index test places a pair of electrodes on a small flat specimen, applies a voltage and drips a conductive solution onto the surface between them. The voltage is reduced in steps until the surface tracks. The result is expressed in volts and describes the resistance of the material surface to tracking under that condition.

The number is comparative, not absolute. It compares materials on a standard specimen under a standard contaminant. It does not tell you what happens on a ribbed housing with dust, humidity and a temperature rise, which is why standards separate the material index from the creepage distance the design must provide.

The test also depends on the specimen condition. A conditioned specimen and a dry specimen can give different results, and the thickness of the specimen matters less than its surface state. Ask which condition the figure was measured under.

The test is also sensitive to the electrode geometry and the dropper rate, which is why the standard fixes both. A result quoted without the standard reference is hard to compare with another supplier's figure, even when both numbers look similar.

Reading the groups

Results are grouped into performance level categories. The highest group covers materials at 600 V and above, the next covers 400 to 599 V, then 250 to 399 V, then 175 to 249 V, and the lowest covers 100 to 174 V. A material in a higher group allows a shorter creepage distance for the same working voltage and pollution degree.

The gap between groups is not linear in cost. Moving from a 250 V material to a 600 V material can change the polymer family, the filler package and the price, and it may change the flame rating and the impact behaviour at the same time.

Read the group together with the working voltage. A design at low voltage with wide spacing may be safe in the lowest group. A compact design at mains voltage may need the higher group simply to fit the creepage into the available wall.

Group letters are sometimes quoted without the standard behind them. Ask for the standard and the test report, then check that the figure applies to the grade and the colour you specified.

Dielectric strength and thickness

Dielectric strength is the voltage at which the material breaks down, divided by the thickness of the specimen. It is usually quoted in kilovolts per millimetre, and the figure is only valid for the thickness and the test condition that produced it.

The relationship with thickness is not linear. A thin specimen often shows a higher strength per millimetre than a thick one, because the probability of a defect in the stressed volume rises with thickness. Designers who multiply a headline kV/mm figure by a wall thickness usually overestimate the withstand of the part.

The test condition matters as well. A short duration test at power frequency, an impulse test and a step test produce different numbers. The specification should name the standard, the waveform and the duration, and the acceptance should be a withstand voltage at the wall thickness used.

Moisture changes the answer. A dry specimen withstands more than a conditioned one, and a housing that absorbs water will move toward the conditioned result in service. For a humid environment, the conditioned figure is the one to use.

From index to creepage and clearance

The material index feeds the creepage distance, not the clearance. Clearance is the shortest path through air and depends on the working voltage, the overvoltage category and the pollution degree. Creepage is the path along the surface and depends on the same factors plus the material group.

That is why a higher group material is worth money in a compact product. It allows a shorter surface path for the same voltage, which can remove a rib, shorten a wall or fit a terminal into a smaller pocket. In a large product the saving may be zero, because the spacing is already generous.

Pollution degree has as much effect as the material. A sealed housing may be assessed at a lower pollution degree than an open one, and that changes the creepage table lookup. If the seal fails in service, the assumption fails with it.

Creepage is measured along the surface, so a rib adds path length without adding width. That is why a moulded groove or a standing rib can solve a spacing problem that a flat wall cannot, provided the rib can be moulded and the tool can release it.

Material families

PA66 and PBT dominate electrical housings, and both can be supplied in grades that reach the higher groups. PC is common where impact and appearance matter, with a tracking index that is usually lower. PPS appears where temperature and chemical resistance matter more than cost, and its tracking behaviour varies widely with the filler package.

The table below is a starting point for a conversation with the material supplier, not a substitute for the data sheet for the exact grade.

A grade that meets a flame rating at 1.5 mm may not meet it at 0.8 mm, and the tracking index can move with the flame package. Treat the two requirements as a pair and check them on the same data sheet, at the same thickness.

Specifying it on the drawing

A drawing that says PA66 does not specify an electrical material. It names a family. The drawing should name the grade, the standard behind the tracking figure, the minimum value, and the thickness at which the flame and electrical ratings apply.

Add the creepage and clearance requirement as a dimension or as a note with the pollution degree and the working voltage. An inspector can measure a distance. An inspector cannot measure an intention.

Finally, control the change. A colour change, a filler change or a switch to a second source can move the tracking index without moving any dimension on the drawing. Treat any material change as an engineering change with a new data sheet and, where the risk justifies it, a new test.

The same discipline applies to incoming inspection. If the drawing names a grade and a minimum index, the goods in check should confirm the grade on the label and on the certificate, not just the colour and the quantity.