There\u2019s a catch. You only get to choose 2 of 3 from design, performance and lifetime. The third is a consequence of the other two choices with consideration of Ohm\u2019s law thermal analogy applied to thermal management of electronics. This is an insight shared by Wendy Luiten, an electronics thermal expert and a design for six sigma master black belt, who has learnt the hard way. Wendy now teaches others, including yours truly.<\/p>\n\n\n\n
The temperature rise over a thermal resistance is proportional to the amount of heat that flows through it. That\u2019s Thermal Ohm\u2019s Law, the thermal equivalent of Ohm\u2019s Law for electrical circuits. It sounds simple, complete and all-encompassing. It is \u2013 it\u2019s quite literally the big picture. But what does this scientific understanding mean for real engineering design?<\/p>\n\n\n\n
In practice the temperature rise is linked to product use and customer expectation regarding usage conditions, lifetime and reliability.<\/p>\n\n\n\n
The heat dissipation links back to customer expectation about performance. Thermal resistance is linked to the design of the product, its size, and aesthetics which give the product its characteristic look and feel.<\/p>\n\n\n\n
So while Thermal Ohm\u2019s Law applies, the power dissipation, reliability and thermal resistance are interdependent. You cannot independently choose performance and featuring *and* use case, lifetime and reliability *and* design. In short, you can only choose two out of three.<\/p>\n\n\n\n
The main reason is that these three aspects of the product\u2019s attributes are worked on separately, by independent groups working in their respective technological silos. No problems will emerge in development and testing as Thermal Ohm\u2019s Law only manifests itself when all three need to act together, typically at the first product integration step when the development budget has been spent and physical hardware exists. The dissipation and thermal resistance having been locked in, and too late to fix the thermal problem, the performance specifications have to be relaxed. At this point the time pressure will be enormous, the situation complex, and compounded by the vested interests of the different groups involved. The underlying cause may never be fully understood.<\/p>\n\n\n\n
Many companies now have incorporated computer simulations to verify the thermal design, and some for over 30 years. Back then simulations took place after the respective departments completed their design and development trajectories, replacing like-for-like hardware prototypes with computer simulations. Simulating a design was cheaper than building a prototype, so the main business driver for adopting simulation was saving cost. Thermal engineers were experimentalists, swapping thermocouples that only show temperature at selected locations, with simulations that showed temperatures everywhere. But there was an unexpected benefit – Insight. Thermocouples don\u2019t show what the air is doing. Computational techniques make it possible to check for conflicts between expectations of performance, usability and design without having to wait for the development stage where hardware would traditionally be available. The thermal models at the time were built at the system level, as it at this integration stage that the problems were identified by prototyping. Without realizing it at the time, in electronics thermal design, the digital twin was being developed in the mid-1990s.<\/p>\n\n\n\n