Two datasheets can quote the same pad, at the same thickness, with numbers that differ by a factor of ten - and both can be honest. The difference is usually the unit: thermal resistance (K/W) versus thermal impedance (°C·cm²/W or °C·in²/W). This page explains what each number means, how to convert between them, and how to compare two TIMs fairly.
The two numbers, defined
| Quantity | Symbol / unit | What it describes | Depends on |
|---|---|---|---|
| Thermal conductivity | k, W/m·K | Property of the bulk material only | Material composition - not geometry |
| Thermal resistance | Rth, K/W (or °C/W) | Resistance of your joint: conduction through the bond line plus both contact interfaces | Thickness, area, pressure, surfaces |
| Thermal impedance | °C·cm²/W (or °C·in²/W) | Area-normalised resistance of the joint: Rth × area | Thickness, pressure, surfaces - but not absolute size |
| Contact resistance | Rc, °C·cm²/W | The two mating surfaces only, excluding bulk conduction | Surface finish, flatness, clamping pressure |
Conductivity is a material property; resistance and impedance are joint measurements. That is why a serious TIM datasheet always states the measurement conditions alongside the number.
How the numbers relate
Z = L / k + 2·Rc [°C·cm²/W], where L is bond-line thickness in cm and k in W/cm·K.
To get joint resistance: Rth = Z / A, with A the footprint area.
Worked example: a 0.5 mm pad with k = 6 W/m·K and contact resistance 0.20 °C·cm²/W per side on an aluminium plate:
Z = 0.05 / 0.6 + 0.40 = 0.483 °C·cm²/W. On a 100 mm × 100 mm footprint (100 cm²), that is Rth = 0.005 K/W - which looks wonderful until you notice the same pad on a 4 cm² die gives Rth = 0.12 K/W. Same material, same impedance, wildly different resistance. This is exactly why absolute-size claims like "0.01 °C/W" are meaningless without the area.
Five traps when comparing datasheets
- Pressure. Soft pads measure far better at high pressure. A value quoted at 100 psi may double at the 20 psi your enclosure actually applies. Always compare at your real mounting pressure.
- Thickness. Impedance scales with bond line. Compare pads at the same thickness, or convert using L/k before comparing.
- ASTM D5470 fixture vs real hardware. D5470 measures through a stacked column with polished isothermal bars. Real housings have bow, roughness and non-uniform pressure - real joints measure worse.
- "Conductivity" quoted instead of impedance. Some datasheets only give k. Convert it, then add an assumed contact term; if the vendor will not state contact resistance at a given pressure, that is a data point in itself.
- °C·in²/W vs °C·cm²/W. 1 °C·in²/W = 6.45 °C·cm²/W. US-sourced datasheets sometimes look better simply because of the unit.
What we quote, and why
Our 180 W/m·K graphene pad is specified as < 0.05 °C·cm²/W total impedance at 50 psi, measured per ASTM D5470, with the pressure stated. The number is deliberately conservative for a soft pad because it includes both interfaces. When a customer shares the real footprint, gap and clamp load, we convert it to the Rth that will appear in their thermal budget - and we publish the conversion, not just the headline.
Quick conversion checklist
| You have | You want | Do this |
|---|---|---|
| k (W/m·K), thickness L (mm) | Bulk impedance | Zbulk = L/10 / (k/100) = 10·L/k °C·cm²/W |
| Z (°C·cm²/W), footprint A (cm²) | Rth (K/W) | R = Z / A |
| Z (°C·in²/W) | Z in metric | Multiply by 6.45 |
| Z at 100 psi | Z at your pressure | Ask for the pressure curve - for soft pads it is not linear |
If you are consolidating a thermal budget, send us the stack: thickness, area, pressure, surface finish. We return the resistance line-item with its assumptions stated, so your budget survives design review.