Home › Blog › Electromechanical & Switchgear › Power Module Thermal Management › Thermal impedance and pad thickness
W/m·K Is Not the Answer: Thermal Impedance, Contact, and the Pad-Thickness Tradeoff (ASTM D5470)
The junction does not care about W/m·K alone; it cares about the whole stack's resistance, which grows with every layer and every extra mil. That single sentence from the parent application page is the cure for most TIM specification errors. Bulk conductivity is one input; the quantity the junction actually sees is thermal impedance: thickness divided by conductivity plus two contact resistances, all pressure-dependent.
ASTM D5470 is the method behind the TDS impedance tables, and reading those tables at your clamping pressure and your thickness, rather than shopping the conductivity column, is the discipline this article walks through, along with the two thickness errors and the slow contact failure the parent page catalogs from the field.
Request a quote for your project →
What ASTM D5470 actually measures
The method reports thermal impedance at defined pressure and thickness: the quantity the stack actually pays. That includes the two contact resistances where the pad meets the module and the sink, which is why impedance changes with clamping pressure and why a comparison made at catalog conditions can invert at yours. The parent page's worked contrast makes the point: a 5 mil pad at the 1.5 W/m·K class regularly beats a 40 mil pad at the 3.0 W/m·K class, and the difference grows as pressure drops.
The same logic explains why a TIM exists at all: it replaces air in the joint, and the contact it makes under real pressure is the performance.
The two thickness errors, and why both cost the junction
Both directions of the pad-thickness error come from choosing thickness in the catalog instead of at the joint. A pad ordered thin to chase the lowest impedance bridges the real, toleranced gap, touches the high spots only, and leaves air over most of the footprint, and air is the enemy the TIM existed to remove.
A filler ordered thick to be safe on a flat, well-clamped joint spends its whole extra thickness as added resistance, the largest resistor in the chain, paid every hour of service.
The fix is the same in both directions: tolerance the gap honestly (component height spread, baseplate bow, fastener torque relaxation), then size thickness to fill the worst case at the available pressure, and no more. Gap Pad TGP grades run from roughly 0.010 to 0.250 inch with softness classes to the ultra-soft grade per TDS; reinforced PROTECT and Sil‑Pad TSP pads live at the few-mil class for flat, clamped faces.
What Stepping Thickness Buys and Costs
| Decision | What it buys | What it costs | Where it is right |
|---|---|---|---|
| Thin reinforced pad (few-mil class) | Least added resistance | Bridges if the gap is real | Flat, well-clamped module-to-sink joints |
| Soft, thicker gap filler | Full-footprint contact across tolerance | Extra mils of path | Multi-height stacks, stamped baseplates |
| Compressible graphite (grounded joints) | Usually wins the thermal comparison | Electrically conductive everywhere | Baseplate-to-sink, grounded spreaders |
| Adhesive film instead of clamping | No dependence on fastener tension | Permanent; prep and rework policy | Fastener-free spreaders, lids, small sinks |
Qualitative reading aids following the ASTM D5470 logic, as framed on the parent application page; the binding data is the impedance table at your pressure on the grade's TDS.
The slow failure: the interface ages out of contact
The stack that measured beautifully at assembly spends five years thermal cycling: fastener torque relaxes, surfaces move, and the interface that depended on yesterday's clamping force loses intimate contact exactly where the heat is.
The parent page's prescription is to design the interface for the stack's whole life: soft, compliant constructions maintain contact as joints move; reinforced pads tolerate re-torque service; and SECURE adhesive films (the 1500KT2 grade reports 53.1 kV/mm per ASTM D149 on its TDS) remove the dependence on hardware tension entirely.
Frame the cycling profile and service interval on the drawing, and treat any TDS impedance value as the fresh-build number it is, with margin for the years after.
Reading a TDS like the junction reads the stack
Three habits keep the comparison honest. Compare at your pressure: impedance tables are pressure-indexed, and the ranking between two grades can flip between rows. Compare at your thickness: a family's headline conductivity says nothing about the gauge your gap demands. And keep path and boundary numbers apart: blanket-insulation conductivity per ASTM C177 answers a different question than joint impedance per ASTM D5470, and the parent page warns against letting the two share a comparison column.
Budget the interface like any other resistor in the chain, and spend thickness only where the gap demands it.
How H-O Products helps
H-O Products is a precision die-cut converter, not a raw-material producer. We convert the pads, fillers, graphite, and films this article discusses: die-cut to the module footprint with fastener clearances, kiss-cut on liner for peel-and-place, laminated spreader-plus-adhesive constructions as one part, with material traceability and lot-code TDS records under our ISO 9001:2015 certified quality management system.
Send the stack drawing with the worst-case gap and clamping force, and engineering checks the operating point against the vendor TDS impedance data at your pressure before anything is quoted.
Frequently asked questions
What do the W/m-K numbers actually tell me?
Bulk conductivity, which is only one input. The junction sees thermal impedance: thickness divided by conductivity plus two contact resistances, all pressure-dependent. ASTM D5470 is the method the TDS impedance tables come from; compare candidates at your clamping pressure and thickness, not by the headline W/m·K. A 5 mil pad at the 1.5 W/m·K class can beat a 40 mil pad at the 3.0 W/m·K class.
How thick should a gap filler be?
Thick enough to fill the worst-case gap at the available clamping force, and no thicker. Tolerance the gap honestly: component height spread, baseplate bow, and fastener torque all open it. Too thin and the pad touches the high spots only, so the effective contact area collapses; too thick and the extra path becomes the largest thermal resistor in the stack. Gap-filler gauges run from roughly 0.010 to 0.250 inch per the maker TDS with softness classes to the ultra-soft grade.
Why did a stack that tested fine at assembly derate years later?
Usually because the interface aged out of contact: fastener torque relaxed over thermal cycles, surfaces moved, and the joint that depended on its original clamping force lost intimate contact where the heat was. Design for the stack's whole life: soft conformable constructions maintain contact as joints move, reinforced pads tolerate re-torque service, and adhesive films remove the dependence on hardware tension. Treat the TDS impedance as a fresh-build number with margin for the years after.
Does this logic change for graphite?
The impedance logic holds, with one addition: graphite is strongly anisotropic, conducting far more in-plane than through-plane, so the comparison depends on which direction your heat travels. A through-path graphite TIM on a grounded joint is compared like any TIM, per ASTM D5470 at pressure; a spreader film is doing a different job, moving heat sideways to real sink area, and needs a destination. Graphite is also electrically conductive in every direction, so the joint's electrical role comes first.
Have the stack reviewed before you lock the pad thickness
Send the stack drawing with the worst-case gap, the clamping force, and the electrical role of the joint, and engineering will check the operating point against the TDS impedance data at your pressure.
Samples typically ship in 3 to 5 business days for common die-cut configurations on materials we keep on hand; standard production runs about two weeks after drawing approval; expedited service is available. MOQ varies by material and part.
Family-owned in Winsted, Connecticut since 1971 · ISO 9001:2015 certified · OEMs worldwide send us a drawing — and get back their part.
Related reading
ApplicationPower Module Thermal ManagementRead →
Sub-applicationServer & Network Thermal InterfaceRead →
GuideTIM Pads, Gap Fillers, Graphite Spreaders, or Aerogel: Choosing Power-Module Thermal MaterialsRead →
GuideHow to Spec Power-Module Thermal Parts: What to Send H-O for a QuoteRead →
IndustryElectromechanical & Switchgear hubRead →
Request a quote for your power-module thermal management part
Send H-O your drawing or describe the part—engineering reviews every RFQ and converts samples on your geometry.
Last updated · Reviewed by H-O Products engineering (ISO 9001:2015 certified converter, Winsted, CT).
References
- ASTM D5470 (thermal impedance of thin thermally conductive solids; the method behind TDS impedance tables, at defined pressure and thickness).
- ASTM D149 (dielectric strength, for joints that must isolate); ASTM D2240 (hardness classes); ASTM C177 (blanket insulation conductivity, the boundary-side method kept out of TIM comparisons).
- UL 94 (flammability classes, per grade).
- Manufacturer technical data sheets for the pad, filler, graphite, spreader, and adhesive-film grades named on the parent application page.
This article explains thermal-impedance reasoning for power-module interfaces using the methods and families cited on the parent application page. The comparisons are qualitative reading aids, not predictions for any design; the binding data is the ASTM D5470 impedance table at your pressure and thickness on each grade's current technical data sheet. All parts are custom manufactured to order; minimum production run quantities vary by material and part.
H-O is a die-cut converter, not a raw-material producer, and does not independently certify materials or finished assemblies.