Dielectric Strength Test: How IEC 60243-1 Measures Breakdown in Solid Insulation

A dielectric strength test pushes a sample of solid insulation to failure with power-frequency AC. You record the voltage where it punctures. You divide by the thickness. You get a number in kV/mm.

That number shows up on most datasheets for plastics, laminates, papers, films, and elastomers. It looks simple. It isn’t. The same material can give very different kV/mm values depending on how thick the sample was, how fast the voltage climbed, what electrodes touched it, and what fluid it sat in.

This article is built around those variables. If you know what moves the number, you can read a datasheet properly, write a test spec that means something, and stop comparing values that were never comparable.

What the Test Covers

IEC 60243-1 is the power-frequency part of the IEC 60243 series. It covers solid insulating materials at 48 Hz to 62 Hz. Part 2 adds the rules for DC. Part 3 covers impulse.

Liquids and gases are out of scope. They only appear as the medium around the sample or as an impregnant. If you need the breakdown voltage of transformer oil, that’s a different test, covered in our oil breakdown voltage test guide.

The US counterpart is ASTM D149. It runs at 60 Hz by default, but allows 25 Hz to 800 Hz. ASTM states that its procedures are all included in the IEC method and the differences are mostly editorial. In practice, the default electrodes and the way results are averaged differ, so the numbers still don’t line up one-to-one. More on that below.

What the Number Is For — and What It Isn’t

This is the part most people skip.

The standard is clear about the purpose. Short-time electric strength is a quality and comparison tool. It catches changes from processing, aging, moisture, or manufacturing drift. It is not a prediction of how the material will behave in your equipment.

Three reasons:

• It’s short-time. Breakdown happens in seconds or minutes. Real insulation lives for decades under much lower stress.
• Discharges are allowed. In most setups, surface discharges at the electrode edges run before breakdown. They change the result. If your design must be discharge-free up to its test voltage, this test won’t tell you the discharge-free strength.
• High kV/mm doesn’t mean long life. A material can puncture at a high field and still fail early in service from heat, partial discharge erosion, or electrochemical attack in damp conditions.

So treat kV/mm as one property among several. For surface behavior, look at comparative tracking index and arc resistance. For long-term discharge damage, look at partial discharge testing of solid insulation.

Breakdown or Flashover?

Before you count a result, decide what actually failed.

• Breakdown is a puncture through the sample. You’ll see a carbonized channel or a hole.
• Flashover is failure of the air or oil around the sample. The current goes around the edge, not through.

Both trip the breaker. Only one is a valid result. Check every specimen after it trips. If there’s no puncture or carbon track, it was flashover. Don’t count it as a breakdown result. Fix the setup: bigger sample, better oil, or a thinner specimen.

One more trap. Breakdown often starts from discharges in the surrounding medium, and the puncture lands just outside the edge of the smaller electrode. The standard describes this as common. It’s still a puncture of the sample, not a flashover.

Variable 1: Thickness

Electric strength drops as the sample gets thicker. Double the thickness and you won’t get double the breakdown voltage. The kV/mm value falls.

That’s why results from different thicknesses aren’t directly comparable. A 0.5 mm film and a 3 mm board of the same resin will give different kV/mm. Neither is wrong.

The standard handles this in a few ways:

• Molded thermosets, thermoplastics, and elastomers are tested at 1.0 mm ± 0.1 mm. If 1 mm thermoset specimens can’t be made, 2.0 mm ± 0.2 mm is the fallback.
• Boards over 3 mm can be machined down on one side to 3 mm ± 0.2 mm when the spec calls for it. The high-voltage electrode goes on the original, unmachined face.
• Thinner machining is allowed if needed to avoid flashover or to stay within the test set’s voltage range.

Then there’s the question of which thickness goes in the math. You have three choices:

1. Nominal thickness (the default unless the spec says otherwise)
2. Average measured thickness of the specimen
3. Thickness measured right next to the puncture point

Option 3 gives the most honest number on a sample with uneven thickness. Whatever you use, write it in the report. A datasheet that doesn’t say is giving you half the story.

Variable 2: Time Under Voltage

This is a big lever. Electric strength drops as voltage stays on longer. A fast ramp gives a higher number than a slow one on the same material.

IEC 60243-1 offers five ways to raise the voltage:

Mode	How it works	Typical use
Short-time (rapid-rise)	Ramp from zero at a steady rate until breakdown, targeting failure in roughly 10 s to 20 s	Quality control, datasheets; fastest and highest values
20 s step-by-step	Start at 40% of the expected short-time breakdown, hold each step 20 s, then step up	Longer stress, lower values than short-time
Slow rate-of-rise (120 s to 240 s)	Slower steady ramp so breakdown lands in 2 to 4 minutes	Same job as a step test, but needs a motor-driven ramp
60 s step-by-step	Same idea as the 20 s test, with 1-minute holds	Longer holds, lower values again
Very slow rate-of-rise (300 s to 600 s)	Slowest ramp, breakdown in 5 to 10 minutes	Longest exposure in the standard

There’s also a proof test option. You apply a set voltage for a set time and record pass or fail, with no breakdown sought.

Two details matter on step tests. First, the time spent ramping between steps counts as part of the hold at the next level. Second, the breakdown voltage on a step test is the highest step the sample survived for the full hold, not the step where it failed.

The short-time test is the default for quality control because it’s fast and repeatable. The longer tests usually give lower values. They also tend to be more useful when you compare one material against another. Keep in mind that none of these modes predicts service life. They are all short-time tests.

Variable 3: Electrodes

Electrode size and shape change the field at the edges. Change the electrodes and you change the result. The standard sets geometry by the form of the material.

Sheets, boards, papers, fabrics, films — tested through the thickness:

• Unequal electrodes (the reference setup): a 25 mm diameter top cylinder about 25 mm high, on a 75 mm diameter bottom cylinder about 15 mm high. Both have 3 mm edge radii. Coaxial within 2 mm.
• Equal electrodes: both 25 mm, but only with a fixture that holds them aligned within 1 mm. Results won’t always match the unequal setup.
• Sphere and plate: a 20 mm sphere on a 25 mm plate with a 2.5 mm edge radius. This is also the arrangement used for elastomers, which was the main addition in the 2013 edition.

Tapes, films, and narrow strips: two 6 mm rods facing each other, with 1 mm edge radii. The top rod weighs 50 g and floats freely, so the pressure on the film is set by gravity, not by the operator.

Tubes:

• Up to 100 mm bore: a 25 mm wide foil band outside, and a tight inner conductor (rod, tube, foil, or packed metal balls) that runs at least 25 mm past each end of the band.
• Over 100 mm bore: a 75 mm foil band outside and a 25 mm foil disk inside.

Molded parts: two 20 mm spheres on a common axis. For shaped parts that won’t sit flat, opposing spheres of 12.5 mm or 20 mm are common.

Filling compounds: two 12.5 mm to 13 mm spheres cast into the compound, 1 mm apart unless the spec says otherwise. Watch for voids between them.

Varnishes and flexible sleeving have their own test standards and follow those. Our insulating varnish test methods page covers the varnish side.

Testing Along the Surface or Along Laminations

Some materials fail between layers, not through them. Laminated boards are the classic case. The standard has three setups for this:

• Parallel plates: a 100 mm × 25 mm strip clamped edge-on between two thick plates. Only for rigid material at least 1.5 mm thick.
• Taper pins: two reamed holes 25 mm apart with 2% taper pins pressed in. Use this when you must stop the failure from running across the surface.
• Parallel cylinders: for very strong boards over 15 mm thick. Two 6 mm rods in drilled holes 10 mm apart.

Keep the electrodes clean and smooth. Pitted electrodes from previous breakdowns can create local field spikes and pull results down, especially on thin film. Stainless steel holds up better.

Variable 4: The Medium Around the Sample

You need something around the electrodes to stop flashover. The choice changes the result.

The usual options are mineral transformer oil, silicone fluid, or synthetic ester. A suitable casting resin can also be used. Very soft elastomers, like silicone rubber, often need to be embedded.

Air works for samples with low breakdown voltage, and for tests at high temperature in an oven. The oven must hold the sample within ±2 K of the set point with circulating air.

Three things to watch:

• Permittivity of the liquid. A liquid with higher permittivity than transformer oil can make the sample look stronger. Swap oil for a higher-permittivity fluid and your kV/mm may go up without the material changing.
• Oil quality cuts both ways. Dirty or wet oil weakens the oil itself. Oddly, that can push the measured strength of the sample up. Either way, the result drifts for reasons that have nothing to do with the material. Keep the oil clean and dry.
• Absorbent materials. Paper and pressboard soak up the medium. The result depends heavily on how they were dried and impregnated. Your procedure must spell this out, or the numbers won’t repeat.

Also check that the fluid doesn’t attack the sample. Swelling during the test invalidates it.

Variable 5: Conditioning

Temperature and moisture both move electric strength. If the material spec gives conditioning rules, follow them. If not, the default is at least 24 hours at 23 °C ± 2 °C and 50% ± 5% relative humidity.

Give the sample and electrodes time to reach test temperature. But don’t leave heat-sensitive materials sitting hot longer than needed.

The Test Set

The source must give a clean sine wave. The peak-to-RMS ratio must stay within ±5% of √2 (1.34 to 1.48) with the sample connected, all the way up to breakdown. So check the waveform under load, not just on open circuit.

For most materials with the standard electrodes, 40 mA output capacity is usually enough. A motor-driven ramp is preferred for the short-time test. ASTM says the same for its slow rate-of-rise method. Hand-cranked variacs give uneven rates, and rate is one of the variables you’re trying to control.

The breaker setting matters too. It defines what counts as “breakdown.” Set it too sensitive and it can trip on leakage, capacitive charging current, or partial discharge pulses. Set it too loose and you get too much burning at the puncture, which also wears the electrodes. ASTM points out that the current setting and response time of the trip circuit noticeably affect results.

If you’re choosing equipment, our hipot testers explained guide covers the hardware side.

How Many Tests, and What to Report

Under IEC 60243-1, the usual practice is five valid breakdowns per material. Take the median, not the mean. If any single value is more than 15% away from the median, run five more and take the median of all ten.

ASTM D149 also calls for five breakdowns unless the material spec says otherwise. But it reports the average, plus a measure of spread, preferably the standard deviation.

The median matters. One weak specimen with a void or a contaminant can drag a mean down hard. The median is less sensitive to one bad sample.

A useful report includes:

• Material, form, and thickness (and which thickness was used in the calculation)
• Electrode type and size
• Surrounding medium and temperature
• Conditioning
• Voltage mode and ramp rate or step time
• Every individual breakdown value, plus the median
• Breakdown location, and any flashovers discarded

Without these, a kV/mm value is just a number.

IEC 60243-1 vs ASTM D149 at a Glance

Item	IEC 60243-1	ASTM D149
Frequency	48 Hz to 62 Hz	60 Hz default, 25 Hz to 800 Hz allowed
Main sheet electrodes	25 mm on 75 mm, 3 mm edge radius	Type 1: 51 mm opposing cylinders, 6.4 mm edge radius
Smaller option	25 mm on 25 mm, with an alignment fixture	Type 2: 25 mm opposing cylinders, 3.2 mm edge radius
Fast test	Short-time, rapid rise	Method A, short-time; 500 V/s is common
Step test	20 s or 60 s steps, start at 40%	Method B, start at 50%, 60 s steps typical (20 s and 300 s alternates)
Slow ramp	120 s to 240 s, and 300 s to 600 s	Method C, slow rate-of-rise
Result	Median of 5 (or 10)	Average of 5, plus standard deviation

Electrode size is one of the biggest differences. A 51 mm electrode stresses a much bigger area than a 25 mm one, and more area means more chance of catching a weak spot. ASTM itself notes that breakdown voltage tends to fall as electrode area grows, and that the effect is stronger on thin specimens. So the same film can read lower on Type 1 electrodes than on IEC’s 25 mm top electrode. The statistics differ too: a median and an average of the same five values won’t always match. If a supplier quotes ASTM and your spec says IEC, ask for retesting or at least agree on which one governs.

Reading a Datasheet Value

Next time you see “dielectric strength: 30 kV/mm,” ask five questions:

1. What thickness? A 1 mm value and a 3 mm value aren’t the same property in practice.
2. What ramp? Short-time values are the highest you’ll see. Step-by-step and slow-ramp values are lower.
3. What electrodes? IEC or ASTM geometry changes the answer.
4. In oil or air? Oil usually allows higher values before flashover.
5. What conditioning? The lab default is 23 °C and 50% RH. Hot or damp service conditions can bring the value down.

If the datasheet can’t answer at least the first three, use the value for rough ranking only. Don’t design clearances or insulation coordination off it.

Quick Reference

Item	Value
Frequency range	48 Hz to 62 Hz
Result	Breakdown voltage ÷ electrode distance, in kV/mm
Reference sheet electrodes	25 mm top / 75 mm bottom, 3 mm edge radius
Tape and film electrodes	6 mm rods, 1 mm edge radius, 50 g upper
Molded specimen thickness	1.0 mm ± 0.1 mm
Thick board reduction	Machine to 3 mm ± 0.2 mm, HV on unmachined face
Default conditioning	24 h minimum at 23 °C, 50% RH
Short-time target	Breakdown in about 10 s to 20 s
Step test start	40% of expected short-time breakdown
Waveform	Peak/RMS within 1.34 to 1.48
Source current	About 40 mA usually adequate
Number of tests	5, median; 10 if any value is over 15% from the median
Valid failure	Puncture through the sample, not flashover

FAQ