March 13, 2020
Heat management remains a critical requirement for electronics, especially in:
Thermal management solutions are successful when they pull heat away from sensitive and/or heat producing components. The best of these systems leverages the optimum heat transfer method — be it conduction, convection and/or radiation.
A heatsink on a central processing unit (CPU) represents one of the most common thermal management solutions. It uses a large surface area and highly conductive materials to passively reduce the temperature of a CPU. Active systems, like fans and liquid cooling, can also be employed.
To select the optimum thermal solution for electronic hardware, engineers start with either finite element analysis (FEA), like Ansys Mechanical, or computational fluid dynamics (CFD), like Ansys Icepak. They utilize these simulation technologies to accurately predict junction and case temperatures across the printed circuit board assembly (PCBA).
But how hot is too hot for these electronic components?
Traditionally, engineers determine the maximum component temperature based on datasheets and derating guidelines. However, engineering leaders are rejecting this outdated approach. Problems with derating tables include:
Designed to be a pre- and post-processor, Sherlock generates thermal and mechanical simulations of electronic hardware.
The pre-processing workflow between Sherlock and Icepak enables the creation of rapid and accurate thermal simulations.
Sherlock reads standard electronic computer-aided design (ECAD) files and then creates part-level geometry — with material properties — to represent the full-featured printed circuit board (PCB). The geometry and material selections are made by utilizing embedded part, package and material libraries.
Since a wide variety of 3D models can be exported from Sherlock into Icepak, users do not need to set up the PCBA or PCB, within Icepak. Icepak users only need to set up heat sources, thermal solutions —including heat sinks and fans — and enclosures. Next, Icepak solves for temperature rises using its solver.
PCBs can be exported as:
Various electronic parts, with lumped thermal conductivities, can also be exported. Even parts without supplier-provided thermal models can be exported such as aluminum electrolytic capacitors.
Sherlock can then import thermal results from Icepak and use its reliability physics engines to predict failure mechanism-specific reliability predictions for a broad range of component technologies. These predictions enable organizations to modernize their derating process and optimize thermal solutions.
For example, should you select a $10 copper heatsink over a $5 aluminum heatsink? Buying 100,000 heatsinks over a year would make this a $500,000 decision. Derating tables will not provide insight into the benefits of each heatsink. However, Sherlock helps users quantify the reduction in warranty costs based on this selection.
Sherlock can also be integrated into a multiphysics simulation — between Icepak and Mechanical — to capture system-level interactions between enclosures and PCBAs. These interactions can cause sudden failures — especially in high power applications such as inverters for electrified drivetrains.
The integration of Sherlock and Icepak demonstrates just one aspect of the Ansys electronic reliability workflow. For more detail, watch the webinar: Ansys Icepak and Sherlock For Temperature Cycling.
We’re here to answer your questions and look forward to speaking with you. A member of our Ansys sales team will contact you shortly.