A thermal pad should compress enough to contact both surfaces across the maximum gap, but not so much that its force exceeds the component, PCB, fastener, or enclosure limit. There is no safe universal compression percentage for every thermal pad.
The familiar “compress it by 20%” rule can be a useful prototype target only when the exact grade’s compression-force curve supports it. Production design must check the full range created by gap and pad tolerances.
Compression ratio formula
If T is the free pad thickness and G is the installed gap:
Compression ratio C = (T − G) / T
Example: a 1.5 mm pad installed in a 1.2 mm gap:
C = (1.5 − 1.2) / 1.5 = 20%
That calculation describes strain. It does not tell you the pressure required to produce it.
Why percentage alone is not enough
Compression pressure depends on:
- Pad formulation and filler loading
- Shore hardness scale and value
- Thickness
- Reinforcement or carrier
- Compression rate and test temperature
- Contact area and surface geometry
- Whether the load is first-cycle or after relaxation
Two pads rated 50 Shore 00 may not generate the same pressure at 20% compression. Hardness is a local indentation test, while your assembly applies compression across an area.
ASTM D575 covers compression-deflection testing for rubber compounds and is useful for comparing stiffness. Ask the supplier for the curve for the exact grade and thickness whenever component load matters.
Define an acceptable compression window
The lower limit should provide continuous contact at the maximum assembly gap. The upper limit should stay below:
- Supplier strain guidance
- Maximum component pressure
- PCB deflection limit
- Fastener and boss load limit
- Enclosure deformation limit
- Any strain at which the pad extrudes, tears, or shifts
Then check tolerance extremes:
Minimum compression = 1 − (maximum gap / minimum pad thickness)
Maximum compression = 1 − (minimum gap / maximum pad thickness)
If minimum compression is negative, the pad can lose contact. If maximum compression requires excessive pressure, the pad can overload the assembly.
The complete thermal pad thickness selection guide includes a worked tolerance example.
Contact pressure versus total force
Do not confuse pressure with force:
Force = pressure × contact area
A pressure acceptable on a small IC may create a large total load across a battery module or wide cold plate. Large-area pads need a clear load path and stiffness analysis.
Also check load distribution. A soft pad spanning mixed component heights may concentrate force on the tallest devices before the shorter ones make contact.
Contact area and flatness must be considered together. As the interface becomes wider, one average gap value is less likely to represent the whole surface. A high spot can carry most of the reaction force while a low corner still has weak contact. Map several points, convert local compression to pressure with the exact grade’s curve, and check how those loads sit relative to fasteners and supports.
If one compression window cannot satisfy every location, increasing the nominal pad thickness is not the only option. Segmenting the pad, using zoned thicknesses, improving the spreader flatness, or changing the support layout may reduce peak force without sacrificing contact.
What happens with too little compression
- Incomplete contact or one-sided witness marks
- Air pockets and higher interface resistance
- Temperature sensitivity to vibration or lid pressure
- Unit-to-unit variation
- Poor contact on shorter components in an array
What happens with too much compression
- Board bow or package stress
- Heatsink lift at an adjacent interface
- Incomplete fastener seating
- Pad extrusion, wrinkling, or tearing
- Higher assembly torque
- Long-term set or shift
Use the diagnostic steps in Thermal Pad Too Thick or Too Thin? if the failure is already present.
How conductivity changes the compression discussion
Higher filler loading can improve apparent conductivity but may also change hardness and force response. Do not trade a known safe pressure window for a higher W/m·K number without system testing.
ASTM D5470 measures thermal impedance under controlled pressure and notes that practical applications do not reproduce the idealized test exactly. Compare impedance at relevant thickness and pressure when data is available.
Jiuju reports high-conductivity thermal pads in the 8–25 W/m·K range with hardness of 40–60 Shore 00, subject to exact-grade confirmation. The company also reports that compression-force curves can be generated to support selection from the customer’s gap and pressure inputs. Request the curve and report conditions before release.
How to read a compression-deflection curve
A useful curve needs more context than two axes. Confirm the grade, free thickness, specimen area, number of layers, platen size, test temperature, compression speed, dwell time, and whether the plot shows a first loading cycle or a later cycle. Viscoelastic pads relax, so force measured immediately after closure can differ from force after a dwell.
Jiuju’s documented equipment list includes a rapid compression tester as well as a Shore 00 durometer and thickness gauge. The durometer checks local indentation hardness. The compression tester applies load over an area and records the response. Use the latter to estimate assembly load.
Suppose the supplier curve reports 120 kPa at your maximum design compression. A 40 mm × 30 mm pad has an area of 1,200 mm², or 0.0012 m²:
Force = 120,000 Pa × 0.0012 m² = 144 N
That is approximately 14.7 kgf, or the force exerted by a 14.7 kg mass under Earth’s gravity. The comparison is only for intuition; the assembly still needs an engineering load limit. If four pads act on the same lid, add their forces and consider their positions. A high pad near a board edge can create bending even when the summed force looks acceptable.
Also compare the curve with production extremes. Use the thickest allowed pad and smallest gap for maximum strain. If the thickness specification is ±10%, a nominal 2.0 mm pad can reach 2.2 mm before assembly. Optional tighter control, such as Jiuju’s reported ±5% capability, may reduce force variation, but only when confirmed for the exact construction and purchase specification.
Ask for the electronic curve data where possible. It is easier to calculate worst cases and compare lots from a table than from a small brochure graph.
A practical validation procedure
- Map minimum and maximum gaps across several assemblies.
- Measure actual pad thickness under a defined gauge force.
- Calculate minimum and maximum compression.
- Convert the grade’s pressure curve to total force over the real area.
- Confirm fastener seating and torque.
- Measure PCB or lid deformation.
- Inspect witness marks on both pad faces.
- Test temperature under controlled load and ambient.
- Repeat after relevant cycling or aging.
For high-risk designs, instrument clamp load or displacement. A temperature pass on day one does not prove the interface will remain mechanically stable.
FAQ
Is 20% compression good for a thermal pad?
It may be, but only if the exact material curve, component load limit, and tolerance stack support it. Treat 20% as a candidate target, not a specification copied across grades.
Does a softer pad need more compression?
Not necessarily. A softer pad may conform at lower pressure. The required strain depends on surface variation and the material curve.
Can I compress a thermal pad to half its thickness?
Do not do so unless the supplier explicitly validates that strain and the assembly can accept the force. Many designs will encounter excessive load or pad deformation well before that point.
How do I measure compression after assembly?
Measure free pad thickness with controlled force and determine installed gap or displacement in the fastened assembly. Avoid using a compressed, removed pad as the only measurement because it can recover or retain set.
Specify a window, not a slogan
The correct answer to “how much should a thermal pad compress?” is a range supported by contact evidence and a force budget. Calculate strain, check pressure, include tolerances, and validate the real assembly.
Jiuju’s thermal pad specifications identify standard grades and Shore 00 hardness values. Send the gap range, area, and pressure limit to request grade-specific selection support.
