I am actually an even-tempered and calm person, but it drives me up the wall when outrageous nonsense is repeatedly spread in discussion forums or YouTube videos. Especially in the field of thermal management, where it seems that everyone is a self-proclaimed expert, misinformation and half-truths that are not only misleading but also technically incorrect are becoming more and more common. This is precisely why it is now time to establish clear facts – through measurements and well-founded analyses.
It must finally be made clear why there can be no uniform thermal conductivity for thermal pads. The actual performance depends on a variety of factors, such as the so-called interface or contact resistance. This describes the thermal resistance at the contact surfaces between the pad, heat sink and components and has an enormous influence on the overall performance. A high W/mK value on the packaging says absolutely nothing if the pad is not in proper contact or the pressure is too low to ensure effective heat transfer. the heat sink
More important than marketing slogans or nicely printed values on the packaging is the effective thermal conductivity of the material at a specific layer thickness, a specific degree of compression and under realistic conditions. This is the only way to assess whether the pad is actually capable of effectively dissipating heat from the hotspots. Without this data, all promises are just empty shells. This is precisely why we are now measuring and showing what really matters. It’s about debunking myths and marketing gobbledygook and finally making it clear that thermal solutions only make sense if they are practical and tailored to the respective application. Facts instead of forum myths – that’s what it’s all about. Let’s finally get started!
Hard thermal pads usually have a rough surface as they cannot be sufficiently adapted to the unevenness of the contact surfaces due to their low compressibility. This uneven contact surface significantly increases the so-called interface or contact resistance, which massively impairs effective heat transfer. While the so-called bulk thermal conductivity of the pad, i.e. the thermal conductivity in the idealized material body without surface resistance, is often advertised with high values of 6 to 14 W/mK, this differs greatly from the actual performance in practice. The Shore value of the pad from page one is also incorrect, it is at least twice as high!
The discrepancy between bulk and effective thermal conductivity is mainly due to the high interface resistance. Hard pads can only compensate poorly for unevenness and air pockets on the contact surfaces, which means that a significant proportion of the thermal energy is accumulated at the interface due to the high thermal resistance and cannot dissipate. The curve of the bulk thermal conductivity over different compression states may appear relatively linear, but the interface resistance prevents good overall performance with hard pads.
Despite theoretically high thermal conductivity values, they are often not convincing in practice. The example of a 1 mm pad advertised with 6 W/mK shown in the following measurement protocol illustrates this problem. With a layer thickness of 1 mm, the pad hardly works, as the interface resistance is extremely high due to the poor contact. Even with a reduction to 950 µm, the bulk thermal conductivity is only just under 3.5 W/mK, which is already a long way from the stated 6 W/mK.
But the real problem lies in the effective thermal conductivity: due to the extremely high contact resistance of almost 99 mm²K/W, the actual performance of the pad is reduced to just 2 to 2.5 W/mK – a value that exposes the marketing claims as pure nonsense. Such results clearly show that the thermal performance of a pad is not determined solely by the stated W/mK values, but by the interaction of material properties, contact behavior and interface resistances. Without taking these aspects into account, manufacturer specifications and promises such as “14 W/mK” only lead to misinterpretations and disappointing results in practice.
The image below of the tested thermal pad after the measuring body has been raised impressively illustrates the problems of hard and poorly adaptable pads. The visible tears and the uneven imprint pattern are a clear indication that the pad was unable to create a complete and even contact surface between the components to be cooled and the heat sink. The tears also indicate that the pad was under mechanical stress, presumably because its hardness meant that it was not flexible enough to adapt to the microstructures and unevenness of the surfaces. Instead of deforming evenly and filling in gaps, the pad partially failed and tore at critical points. Such defects act like thermal barriers as they either leave air pockets or completely lose contact with the heat sink.
The poor impression also confirms that the thermal interface resistance must be exceptionally high. The few clear contact areas on the pad show that the contact pressure was locally insufficient to allow effective heat transfer. At the same time, some areas were probably over-compressed, leading to the tears. This unevenness is a typical problem of hard pads, as they do not distribute the pressure evenly and thus greatly reduce thermal efficiency.
In practice, such a condition leads to a drastic deterioration in cooling performance. Even if the pad had a theoretically high bulk thermal conductivity, this is completely undermined by the lack of contact and the high interface resistances. The result is poor temperatures, which are often wrongly attributed to the thermal conductivity of the material or the cooling design, although the cause lies in the pad’s mismatch.
Note: A hard thermal pad with a nominal thickness of 1 mm is typically subject to manufacturing tolerances in the range of ±0.1 mm to ±0.15 mm. This means that the actual thickness of such a pad can range from approximately 0.85 mm to 1.15 mm, depending on the precision of manufacture and the quality requirements of the manufacturer. These tolerances are all the more critical for hard pads, as they can lead to uneven pressure distribution and poor heat transfer in practice, especially for components with sensitive height differences or uneven surfaces.
These tolerances can be particularly problematic with hard pads, as their low compressibility and adaptability are barely able to compensate for manufacturing variations. A pad that is 1.15 mm thick, for example, requires more contact pressure to compress to the required thickness, which is not achievable in many cases. At the same time, a pad that is too thin (e.g. 0.85 mm) may not make full contact with both surfaces, creating air gaps that drastically increase thermal resistance.
Pages: 1 - Introduction and Basics of Pad Selection 2 - We are delving deeper (and thicker). 3 - Measurement: Thin and Hard Pads 4 - Measurement: Thin and Soft Pads 5 - Measurement: Thickness, Ultra-Soft Pads 6 - Measurement: Thermal Putty, Pad Stacks, Recommendation, and Conclusion
Top! Thank you for the revealing article
Now only one article is missing about how to use the necessary layer strengths on a graphics card u./o. Similar.
Igor had already shown one possibility when saving a graphics card. Free from the memory: - Card and cooler - Clean the area from the cooler of the on chip once laying down with "Tesafilm" (as a WLP -layer fake) - build up the card - split dimensions between cooler and cooling components (RAM, MOSFETS, ... but not the GPU chip) using a gaps/sensor teaching - determine the next strength (Soft) Take the thermal pad
Thanks! The keyword was the gapshive apprenticeship/sensor teaching. For the pads there are also purposes away from graphics cards. A defective SSD, if you take the wrong (too thick and too hard) pad size when replacing the mobo equipment, is just as annoying ...
My conclusion: The best freaking with different layer strengths and hardening of the pads is best way to get out of the way with a good putty.
But even with very soft putty, you get to the point at some point where too much contact forces are necessary for the deformation of the putty. And exactly when you use the putty too large.
Read all comments at Igor´slab Community →
Editor-in-chief and name-giver of igor'sLAB as the content successor of Tom's Hardware Germany, whose license was returned in June 2019 in order to better meet the qualitative demands of web content and challenges of new media such as YouTube with its own channel.
Computer nerd since 1983, audio freak since 1979 and pretty much open to anything with a plug or battery for over 50 years.
Follow Igor: YouTube Facebook Instagram Twitter
cpu and heat sink We use technologies such as cookies to save and/or access device information. We do this to improve the surfing experience and to display personalized advertising. If you agree to this technologies, we can process data such as surfing behavior or unique IDs on this website. If you do not give or withdraw your consent, certain functions can be impaired.