The specimen decides the number almost as much as the material does. A tracking test measures how a surface behaves under a contaminated film, and nearly every rule about how you cut, clean, and condition that surface exists to stop the specimen itself from becoming the variable.
Get the prep wrong and the result is not slightly off — it can land in the wrong material group. This covers specimen preparation for the two tracking standards, the drop test in IEC 60112 and the inclined plane test in IEC 60587, stage by stage, with the failure each rule is there to prevent. For the tests themselves, see the comparative tracking index pillar and the IEC 60587 inclined plane page.
The headline difference to hold in mind: these two tests want almost opposite things from a surface, because one drips contaminant onto a flat plate and the other runs it down an incline.
Table of Contents
Dimensions and flatness
IEC 60112. Any roughly flat surface works, as long as it is big enough that contaminant does not run off the edge during the test. Flat areas of at least 20 mm × 20 mm are recommended; smaller sizes are allowed only if no electrolyte is lost, such as 15 mm × 15 mm multipurpose specimens. Separate specimens per test are preferred. If you run several tests on one piece, keep the sites far enough apart that splash and fumes from one do not contaminate the next.
IEC 60587. Much larger — at least 50 mm × 120 mm — because the contaminant has to flow down a channel between electrodes 50 mm apart, not just pool between tips 4 mm apart. The specimen is also mechanically processed with boreholes so the electrodes can be bolted on.
The error it prevents: electrolyte running off the edge. If the film escapes instead of accumulating, the surface never sees the contamination the test depends on, and the material scores high for the wrong reason. On the drop test, a foil detector around the glass support is the usual way to catch electrolyte loss.
Thickness and stacking
IEC 60112. Minimum 3 mm. If a single piece is thinner, stack pieces to reach it. Stacking is explicitly allowed for a reason: thinner specimens transmit more heat into the glass support underneath, which changes how the contaminant film dries between drops. Values from specimens under 3 mm are not comparable to values from thicker ones.
IEC 60587. Preferred thickness is 6 mm. Other thicknesses are permitted but must be stated in the report.
The error it prevents: a hidden heat-sink effect. The drying and re-wetting cycle on the surface is what drives tracking. A thin specimen dumps heat to the support and dries differently, so its number does not mean the same thing as a thicker one’s. Match thickness — or stack — before comparing results.
Surface state — where the two tests disagree
This is the one to read twice, because the two standards pull in opposite directions.
IEC 60112 wants the surface pristine. Specimens should be nominally smooth, untextured, and free of scratches, blemishes, and impurities unless the product standard says otherwise. Surface features add scatter to the results. Tests are run on clean surfaces, and the cleaning method is reported — but cleaning must not swell, soften, or abrade the material, because that damage skews the number too. Roughening here is a defect.
IEC 60587 will make you roughen the surface on purpose. Good wettability is a stated prerequisite — if the contaminant beads instead of spreading, the test does not work. So the standard sanctions controlled abrasion: grind lightly with a fine aluminium-oxide or zirconia-alumina abrasive under water until the whole surface wets, then rinse. Roughening here is a fix. The alternative is to temporarily raise the contaminant flow rate until the surface wets before the voltage goes on. Either way, any grinding is noted in the report.
Why the reversal? The drop test drops discrete beads onto a flat plate; a smooth surface keeps the geometry repeatable. The inclined plane test needs a continuous film running down the face, and a surface that will not wet breaks the whole method. Same property — surface condition — but the delivery mechanism flips what “good prep” means.
The error it prevents: on 60112, surface features and cleaning damage inflate scatter and can shift the result. On 60587, a non-wetting surface produces an invalid test — no continuous film, no meaningful scintillation.
Conditioning
Both standards condition the same way, near enough: at least 24 hours at 23 °C and 50% ± 10% relative humidity. IEC 60112 allows ±5 K on temperature; IEC 60587 holds a tighter ±2 °C.
The error it prevents: moisture state drift. A material’s surface conductivity depends on how much water it holds, and some polymers — nylons especially — absorb enough to move the result. Fixed humidity and time put every specimen in the same state so the number reflects the material, not the weather in the lab that week.
Orientation and fabrication
Anisotropy. Where the material has a grain — laminates, fiber-filled compounds, parts with a strong flow direction — the tracking result depends on which way the electrodes face relative to that feature. Both standards say to test both orientations and report the one giving the lower value, unless something specifies otherwise.
Fabrication route. IEC 60112 warns that fabrication conditions and flow direction change PTI and CTI performance. When you cannot cut a specimen from an actual part and use a molded plaque instead, match the fabrication process as closely as possible — a plaque made differently from the part can give a different number.
The error it prevents: reporting a value that does not represent the part. Test only the easy orientation and you may miss the weak direction that fails in service. Test a plaque made by a different process and you characterize the plaque, not the product.
Mounting (IEC 60587)
The inclined plane test adds a mounting step the drop test does not have. The specimen sits at 45°, contaminant-side down, electrodes 50 mm apart, bolted through the boreholes. Two things have to be right: no gap between the electrode edges and the specimen surface, and no deformation of that surface under the clamping. A fresh filter-paper stack goes on for every test. If the specimen is not self-supporting, it rides on an insulating, heat-resistant support such as PTFE, chosen so it does not block heat dissipation from the back.
The error it prevents: uneven contact and distorted geometry. A gap or a bowed surface changes where and how the scintillations form, which changes the result.
Quick comparison
| Prep parameter | IEC 60112 (drop) | IEC 60587 (inclined plane) |
|---|---|---|
| Minimum size | ~20 × 20 mm | 50 × 120 mm |
| Preferred thickness | ≥ 3 mm (stacking allowed) | 6 mm |
| Surface goal | Smooth, clean, undamaged | Must wet — abrade if it won’t |
| Roughening | Avoid — it’s a defect | Sanctioned fix for non-wetting |
| Conditioning | 24 h, 23 °C ±5 K, 50% ±10% RH | 24 h, 23 °C ±2 °C, 50% ±10% RH |
| Electrode fixing | Placed on surface | Bolted through boreholes |
| Mounting | Flat, horizontal | 45°, contaminant-side down |
FAQ
What size specimen does the CTI test need? For the IEC 60112 drop test, a flat area of at least 20 mm × 20 mm is recommended — small enough that contaminant does not run off the edge. The IEC 60587 inclined plane test needs much larger specimens, at least 50 mm × 120 mm, to give the contaminant a channel to flow down.
Why does specimen thickness matter? Thin specimens transmit more heat into the support beneath them, which changes how the contaminant film dries between drops. IEC 60112 sets a 3 mm minimum for this reason and allows stacking to reach it. Values from thinner specimens are not comparable to thicker ones.
Should I sand or polish the specimen before testing? It depends on the standard. IEC 60112 wants a smooth, undamaged surface and treats abrasion as a defect. IEC 60587 requires the surface to wet, and if it does not, it allows controlled grinding under water until it does. Any grinding is recorded in the report.
How should specimens be conditioned? At least 24 hours at 23 °C and 50% ± 10% relative humidity, under both standards. This fixes the moisture state so the result reflects the material rather than ambient humidity.
Can I test a finished part instead of a plaque? Yes, if you can get a flat, large-enough, thick-enough area. Where you cannot, a molded plaque of the same material may be used — but match the fabrication process, because a plaque made differently can give a different CTI than the part.
Which way should the electrodes face on an anisotropic material? Test both the direction of the grain or flow and the direction across it, then report the orientation that gives the lower value, unless a specification says otherwise. This ensures the weak direction is not missed.