How to Choose a Thermally Conductive Potting Compound: A 4-Question Framework
The narrowest gap in the cavity was 0.4 mm. The compound on the line was a 12,000 cps grade. At that viscosity the material bridges the gap, cures on top, and leaves an air pocket underneath — and the humidity test finds it.
The material had been chosen on thermal conductivity alone. That is the one number most buyers compare when they select a thermally conductive potting compound. Conductivity was in spec. It was never the issue.
The opening is a composite of recurring field failure modes, not one customer’s case.
You buy a potting compound on four conditions. Conductivity is one of them.
01 The four questions that set your potting compound grade
Ask them in this order. Stop as soon as one question rules a material out.
Question 1 — How much heat, and through how much area?
This sets your thermal grade. Steady-state conduction through the compound gives the minimum conductivity you can get away with:
SI units: P in W · t in m · A in m² · ΔT in K
What each symbol means, and where engineers get it wrong:
P = the heat the device actually dissipates, in watts. That means loss power, not nameplate rating. A 100 W converter running at 95% efficiency puts 5 W into the potting, not 100 W. Engineers most often plug in the rated output here.
t = compound thickness between the heat source and the cooling surface, in metres.
A = the heat-spreading area under the device, the projected footprint in m². Not the outer housing surface.
ΔT = the temperature rise you accept across the compound, in kelvin. A difference of 1 K equals 1 °C, so a Celsius budget reads straight across.
Units must stay consistent. In shop-floor units the same formula reads:
Same result — only the unit set changes
SI: (5 × 0.002) / (0.001 × 20) = 0.5. Shop-floor: 10 × 5 × 2 / (10 × 20) = 0.5.
→ kmin = 0.5 W/(m·K). Our 1.0 grade (≥ 0.9 W/m·K) clears the 1.5× target of 0.75 with margin. Step up a grade when the ΔT budget is uncertain, when the footprint is tight, or when the assembly will run hot — this number climbs fast when power is high and the area is small.
Question 2 — What is the narrowest gap?
This sets your viscosity, and it is the question people skip. A 0.4 mm gap will not take a 12,000 cps compound. The gap table below is built for this question.
Question 3 — Will it ever need to come apart?
This sets your chemistry. Addition-cure silicone is elastomeric and can be cut out for rework. Epoxy is rigid and generally cannot. If your customer mentions field repair, epoxy is the wrong answer.
Question 4 — Can the line apply heat?
This sets your cure system. Addition-cure silicone runs fast at 80 °C. If the line can only do room temperature, you need a condensation-cure grade, and you need to plan curing time into the cycle.
Every product in our catalogue is indexed by these four answers. Send us your four answers and we will shortlist a grade and tell you what to verify on your line.
02 Thermal grades: what actually changes
Values are nominal. Refer to the TDS for the tolerance on your part number.
| Grade | Part no. | Conductivity (W/m·K) | Mixed viscosity (cps) | Hardness · Density (g/cm³) |
|---|---|---|---|---|
| 1.0 | FD503-TG101 | ≥ 0.9 | 2,500 ± 1,000 | 45–55 Shore A · 1.72 ± 0.05 |
| 1.5 | FD503-TG102 / TG15 | ≥ 1.5 | 4,000–6,000 | 30–40 Shore A · 2.15 ± 0.1 |
| 2.0 | FD503-TG20 | 2.0 ± 0.2 | 3,000–5,000 | 20–30 Shore A · 2.73 ± 0.1 |
| 3.0 | FD503-TG301 | 3.0 ± 0.3 | 5,500–7,500 | 45–65 Shore 00 · 3.0 ± 0.1 |
| 3.0 | FD503-TG30 | ≥ 3.0 | 8,000–15,000 | 40–60 Shore A · 3.0 ± 0.1 |
| 4.0 | FD503-TG40 / TG40H | 4.0 ± 0.4 | 10,000–15,000 | 60–70 Shore A · 3.0 ± 0.1 |
Two things stand out. Viscosity does not track conductivity: the 2.0 grade flows thinner than the 1.5 grade, and the 4.0 grade is only about twice as thick as the 2.0. And the 4.0 grade shares its density with the 3.0 grade (3.0 g/cm³) while conducting more heat — filler type and particle packing move the conductivity, not only how much filler goes in. Price a grade against its TDS, not against the number in the product name.
Density still shows where the cost sits. Grade 1.0 carries 1.72 g/cm³ of filled system; grade 3.0 carries 3.0. That is roughly 75% more filler mass per litre of product, and filler is the expensive part. That is why the gap between a 1.0 and a 3.0 thermally conductive potting compound is mostly a filler cost, not a formulation secret.
03 The table nobody publishes: minimum gap vs viscosity
This is the row that took the line down in the opening example. For a thermally conductive potting compound, viscosity decides whether the material ever reaches the bottom of the cavity — conductivity only matters after it gets there. The ranges below are engineering experience values — a shortlist, not a certified dataset. We are running our own gap-flow tests on video to replace them with measured values; ask us for the latest test record.
| Mixed viscosity (cps) | Flow class | Minimum gap it fills | Our part numbers |
|---|---|---|---|
| ≤ 1,500 | Ultra-low / penetrating | 0.10 – 0.30 mm | FD503-TG60S (600–1,200) |
| 1,500 – 3,000 | Low | 0.20 – 0.50 mm | FD503-TG60E (1,500–2,500) |
| 3,000 – 6,000 | Medium | 0.40 – 0.80 mm | FD503-TG20 (3,000–5,000), TG102 / TG15 (4,000–6,000) |
| 6,000 – 15,000 | High | 0.80 – 2.00 mm | FD503-TG30 (8,000–15,000), TG40 / TG40H (10,000–15,000) |
Band boundaries are for shortlisting. A part whose viscosity range straddles a boundary should be verified on your own line — that applies to FD503-TG101 (1,500–3,500), FD503-TG152 (800–1,800) and FD503-TG301 (5,500–7,500).
FD503-TG152 is a 6:1 condensation-cure product, not a 1:1 addition-cure grade: mixed viscosity 800–1,800 cps, hardness 35–45 Shore A, conductivity ≥ 0.5 W/m·K.
04 Before you commit: three checks
Incoming. Check the TDS revision against the delivered drum. Confirm mix ratio by weight, not by eye. A 1:1 product mixed at 1:1.15 will cure, but not to spec.
In process. Measure the actual fill time on one sample unit. If it runs longer than the table above suggests, your cavity has a shadow area and you need a lower viscosity or vacuum assist.
After cure. Cut one unit open. Look for voids and for an unfilled gap bottom. Do this once per batch. It takes ten minutes, and it is the check that catches the failure mode this article opened with.
05 One process note that costs whole batches
If you mix with a shared tool, clean it between every batch. Addition-cure silicone is inhibited by trace contamination — sulphur, tin, some amines, and residue from a previous batch’s Part B. A dirty paddle can stop a whole batch from curing, and the failure shows up hours later at the end of the line.
A dedicated process guide covers this in detail; ask us for it. For now: one tool per material, cleaned and dried.
06 Where to buy a thermally conductive potting compound
Buy a technical material from a manufacturer, not a listing. Three things to ask for: a downloadable TDS with tolerances, a Certificate of Analysis (COA, the per-batch test report) for every lot you receive, and a technical contact who answers within 24 hours. Price matters, but it is not the first question.
If your application is outside the general-purpose cases above — inverters, energy storage, outdoor electronics — start from the
application scenarios and selection approach instead, then come back to the four questions.
Related reading
- Inverter potting compound selection: conductivity, processability and long-term reliability
- Two-component or one-component potting with FD-503 addition-cure silicone
- NEV penetration past 50% and what it changes in potting demand
- Potting compound application scenarios and selection approach for 2026
- The future of thermally conductive materials
07 Appendix: units used in this article
| Quantity | Written here | Equivalent | Note |
|---|---|---|---|
| Thermal conductivity k | W/(m·K) | W/m·K, W/(m·°C) | All three mean the same thing; the parenthesised form avoids being read as W·K/m |
| Power P | W | — | Must be the actual dissipated loss, not the rated output |
| Thickness t | mm | 1 mm = 0.001 m | The SI form takes metres, the shop-floor form takes mm directly |
| Area A | cm² | 1 cm² = 0.0001 m² | Projected footprint under the device, not housing surface |
| Temperature rise ΔT | K | °C | A difference of 1 K equals 1 °C |
| Viscosity | cps | mPa·s | 1 cps = 1 mPa·s. Our TDS uses mPa·s; this article uses cps — same value |
| Density | g/cm³ | — | As in our TDS |
| Hardness | Shore A / Shore 00 | — | As in our TDS |
Send us your cavity drawing and gap measurement
We will recommend a grade from the table above — and tell you what to verify on your own line. Sample kits: 100 g / 1 kg / 20 kg.
Zhuhai Double Carbon Technology Co., Ltd. · Brand: Double Carbon · www.dcadhesive.cn
