Never miss a story from Computational Fluid Dynamics. Subscribe for in-depth analysis and articles.
We have come a long way, considering electronic gadget size and battery life. Look at the cellular phones; the first ever portable cell phone designed by Motorola lasted only 30 minutes or less and weighed about 3 pounds. These phones were called ‘bricks’ for obvious reasons. Some of today’s smartphones are faster and more powerful than PCs. The miniaturization of electronic goods with the need for high-power density adds immense pressure on electronic cooling. Electronics cooling technology has been in place since the 1960s, but as customers aspire for high-power electronic gadgets, new thermal management techniques become prerequisites for accommodating these needs. There is a lot to learn from the past, apply those learnings with necessary modifications to the present, and predict the needs of future cooling systems.
The advent of CFD for electronics cooling started only during the late 1980s, approximately 30 years after CFD found its place in aerodynamic applications. The first workstation installed for CFD studies on electronics cooling used Unix. For using this platform, it was an absolute necessity for thermal engineers to have a core understanding of fluid dynamics and heat transfer. The geometry laid out in green and black wireframes, took days and weeks to build. Each geometry would be represented as a set of cuboids consisting of only rectangular blocks and triangular prisms. During those days, a 20,000-cell grid would take weeks of number crunching, and after all those calculations, there were high chances for the solution to diverge.
A few changes came along as time passed, including improvements to the graphical interface, MCAD, and ECAD import, and the computational time was reduced significantly. But the geometry modeling and discretization aspects of CFD solutions remained stagnant, and there haven’t been any significant advancements in the last ten years. The growth curve often declines when only certain aspects of a particular application are addressed or improved, and these enhancements do not fit into the crude system.
By starting from scratch, software such as Cadence Celsius EC Solver has drastically changed the electronics thermal simulation landscape. This software has become intelligent. Instead of relying on the engineer to correctly form collections of generic cuboids into the items to be modeled and get the gridding right, Celsius EC Solver has specific items for specific functions. For example, it knows that a heatsink is finned, conducting solid that requires airflow for adequate cooling, and sensors control fans and PCBs come with layers, traces, and percent copper regardless of the units used.
Gridding is performed automatically and intelligently rather than just by keypoint. It is an intelligently generated grid reflecting decades of industry experience to capture physics and geometry precisely. It is also optimized for your model and your computer's computing capacity. A fast, distributed memory, parallel processing algorithm can use low-cost multicore and multi-processor computer systems to reduce compute times. Models with more than 100,000,000 grid cells are nothing more than an overnight run.
The development of current technology, viewed from 1989, is just as revolutionary as that of CFD in electronics cooling. The time necessary to perform thermal analysis has been drastically reduced, in most cases, by at least half. But the true revolution combines these new time savings with the fact that a thermal engineer needs no longer be a specialist. A generalist can analyze most issues effectively by leveraging MCAD and ECAD data and using intelligent software. It does not mean that thermal engineers are no longer necessary or that anybody can do thermal design. Thermal analysis using CFD is a function that still requires the engineering mind's knowledge, training, and expertise.
What we do today is expected to define our future! The department of energy in the US has come up with a new cooling technique that uses light emitting diode (LED). The density of photons is dropped when electricity in the LED flows in a direction opposite to normal. This work carried out by the team is expected to bring about advancement toward nanophotonic cooling. This cooling method has an excellent opportunity for on-chip cooling in the future.
1. Cooling Electronics of the Future. Basic Energy Sciences.
2. Ross, Marie. Electronics Cooling CFD Trends.
Learn more about our other Cadence CFD product - Fidelity CFD, by clicking the button below -