Discussion of Thermal Management

Introduction

The electronics industry has made tremendous progress in recent years to become an integrated part of our daily lives. Progress has come on two main fronts: increased functionality in a single device and miniaturization of devices. Both of these developments have increased power and energy required or produced, thus increasing heat and the need for thermal management within devices.

There are many different methods to remove or transfer heat energy. Desktop PC’s tend to use aluminum heat sinks with fans. Notebooks utilize heat pipes and thermal interface materials to connect a heat source to a metal chassis in conjunction with fans. More recently, synthetic, and natural graphite has been used in enclosed environments such as smartphones, tablets, hyper-slim ultrabooks and other electronics-oriented devices where fans are limited or cannot be used due to space, environment or noise constraints. Boyd’s discussion of Thermal Management will include various thermal applications within a host of diverse devices and end markets.

Thermal Energy Transfer

Heat is associated with the inefficient transfer of potential and kinetic energy within a system. Thermal energy always moves from a warmer system to a colder system. The transfer or dispersion of heat can occur by three main mechanisms: conduction, convection, or radiation. CONDUCTION is the flow of heat through solids and/or liquids via the vibration and collision of molecules and free electrons. This transmission method is most common in electronic devices regardless of having a sealed or ventilated system. The formula to calculate the conductivity gradient for a given system is: q = – kA (Δ T/Δ n) Where Δ T/Δ n is the temperature gradient in the direction of area A, and k is the thermal conductivity constant of the material obtained by experimentation in W/m-k. CONVECTION is the flow of heat via currents within a fluid (liquid or gas). Convection is the displacement of volumes of a substance in a liquid or gaseous phase. This mechanism is common in large devices with enough space for air or liquid circulation. The formula for convection: q = hA (T_s – T_∞) Where h is the convective heat transfer coefficient, A is the area implied in the heat transfer process, T_s is for the temperature of the system and T_∞ is a reference temperature. RADIATION is heat transfer by electromagnetic waves or photons; it does not need a propagating medium. The energy transferred by radiation moves at the speed of light. For example, the heat radiated from the Sun can be exchanged between the solar surface and the Earth’s surface without heating the transitional space. The formula to calculate the amount of heat transferred by radiation is: q = e σ A [(ΔT)⁴] Where q is the heat transferred by radiation, E is the emissivity of the system, σ is the constant of Stephan-Boltzmann (5.6697*10⁻⁸ W/m².K⁴), A is the area involved in the heat transfer by radiation, and (ΔT)⁴ is the difference of temperature between two systems to the fourth or higher power.

An example of this is the read/write operation of a typical hard disk drive that undergoes rapid rotary actuation of the arm assembly as it moves from track to track over the disk media during normal read/write operations. Each hard start & stop action is like a miniature impact hammer hitting the structure and energizing all its internal resonances. This is problematic to the extent that there is not sufficient damping in the system to quickly dissipate this unwanted motion of the read/write head. Ultimately, off-track errors can occur slowing down the performance of the device.

Elaborating on this example further, a typical hard disk drive is subject to a collection of excitation sources: the rotating spindle motor that spins disk platters to 7500 or 10000 RPM, the rotary actuated voice coil motor that pivots the arm actuator assembly, the bearing effects of either the arm pivot or spindle bearing that create unwanted disturbances, and the air induced excitation from turbulent airflow from the spinning disks pushing air over the arm / suspension components. The goal is to control these excitation forces at the source level through various design choices. For example, a significant improvement in idle noise created by the drive was achieved once the drive industry changed to more precise fluid dynamic bearings. Better quality components manufactured to higher quality standards reduced variation by tightening component tolerances further helping to reduce excitation levels. Air induced vibration is a by-product of the high speed drives made today, but even this source can be controlled through the use of air straightening devices that help to minimize turbulent air flow, thus reducing this source of broadband excitation to the disk platters and actuator.

In general, strategies for minimizing excitation source levels involve such things as use of light weight components to reduce force levels, minimizing unbalance and misalignment between components, and more precise manufacturing methods that remove unwanted variation. The reduction of reciprocating loads can be achieved by reducing the mass of moving components or the use of inertial counter balances. For geared components, selection of high contact ratios (>2.0), proper lubrication, selection of gear materials, tooth profile and surface finish, and shaft alignment are all factors influencing good gear design and operation. Other methods involve the modification of the actual operating profile whereby sacrifices in speed or power are made for the benefit of better NVH characteristics (i.e. “quiet mode” of a cooling fan that runs at a slower speed often actively controlled to control cooling demand, or an automotive air conditioner that takes longer to cool because of less powerful components, or a hard drive that decelerates slowly to a stop minimizing excitation levels at the expense of longer seek times).

Computational models, enabled via advanced Software, allows for analysis of the thermal characteristics of a system, including computational fluid design (CFD), computational heat transfer (CHT), power management, circuit design and other electronic design automations (EDA). Good software capability maximizes the value of a heat sink or thermal module by confirming, through design, the most effective heat transfer path before device design is locked down.

Substrate: having been developed specially to enhance the thermal handling capability of an electronic component, this sub-segment includes two phase heat spreaders, graphite heat spreaders, diamond heat spreaders and silicone based heat spreaders. Recently, graphite has been chosen as a lead substrate product in smartphones and other sealed system devices due to its thin profile, high in-plane thermal conductivity and low weight.

Where is Thermal Management Most Commonly Needed?

Major consumers of thermal management products reside in industries like consumer and enterprise electronics, telecommunications, automotive, medical devices and industrial applications. The most common thermal management solution is the heat sink (aluminum or copper in Figure 2) or cooler module (heat pipe in Figure 3) laminated with interface material such as a gap pad. If the design permits, a fan is installed for accelerated heat dissipation.

Sealed system thermal management is an entirely different topic when discussing heat dissipation due to system design. In an open system, we can easily use air circulation to exchange heat to the environment. Sealed systems, however, typically have no room for tall heat sinks and inherently do not allow air circulation within the device. The most common thermal management technique in a sealed system uses a thermal spreader or shield, typically made of graphite, to either increase the available surface area with which to spread a hot spot, or, in conjunction with an air gap, shield a sensitive component from a heat source.

Graphite, due to its layered molecular structure, is anisotropic, meaning it spreads heat in plane much more quickly than through plane. As a result, graphite eliminates hot spots very well compared to isotropic materials, such as metal, which move heat in all directions uniformly, as illustrated in Figure 4.

Graphite has two major functions: Thermal Spreading: when in contact with a heat source, graphite spreads heat across its surface area, increasing area available to dissipate heat, and cooling down a heat source. Thermal Shielding: when designed with an air gap between the graphite and the heat source, graphite can uniformly distribute the heat of a hot spot and reduce the surface temperature of the opposite side, relative to the heat source, of a device.

Thermal Management Drivers

As a result of the widespread introduction of microelectronics across market segments, together with the increasing demands on functionality and reliability, thermal management has become an important issue in almost every branch of the technology world, including professional and consumer electronics systems as well as automotive electronics, set top / home gateway boxes, LED lamps and medical devices.

From a thermal management perspective, there is a significant price to pay for the increase in device functionality. Device operating frequency and gate counts are increasing rapidly, dissipating greater amounts of power as heat. The buildup of excessive heat is a major cause of failure in electronic systems. Electronics industry reports indicate 55% of all failures are caused by “temperature”, and further recommend that decreasing electronic component temperature by 10°C could on average double the lifespan of a device (refer to Arrhenius’ Law of Chemical Activity in Illinois Capacitor Inc. report on life calculator).

Since the generation of heat is representative of inefficient use of energy, it stands to reason that several global energy conservation and power management standard initiatives, such as the Environmental Protection Agency’s (EPA) Energy Star, the European Union’s (EU) Blue Angel and the Advanced Configuration and Power Interface (ACPI) in the computer industry, benchmark thermal performance as an illustration of energy efficiency.

As electronic devices have increased in power, consumers have seen media coverage featuring devices catching fire and discussing theories of thermal runaway. Thermal runaway is chain of cyclical reactions of temperature rise that build upon one another, increasing device operating temperature to critical levels, often leading to device shut down from overheating. In rare cases, thermal runaway can lead to ignition but, thus far, there is no direct evidence to indicate overheating causing electronic device ignition. There may not be a thermal management mechanism failure in these situations, but consumer perception of device overheating causing ignition is something we work to minimize by designing appropriate heat management into devices.