The trend for reducing the size of power electronics systems requires smaller footprints for power Metal Oxide Semiconductor Field-Effect Transistor (MOSFET) and Insulated Gate Bipolar Transistor (IGBT) modules. This has resulted in higher power dissipation densities for the IGBT die, as well as the module, due to a higher packing density of the die. Also, increases in switching frequencies and voltage ratings of IGBTs result in higher power dissipation at the die level. Some of the die power losses have been offset by advances in both MOSFET and IGBT chip design, but the cooling capabilities of present modules limit the device performance.

The Navy’s next-generation multi-mission destroyers will reduce the size, weight, and cost of its power electronic modules. This requires components that have higher current densities, higher switching frequencies, and higher operating temperatures than current generation electrical systems. However, power electronics packages available today are not designed to meet ship-board environmental and operational demands due to an unacceptably high package thermal resistance. Additionally, commercial component suppliers are generally averse to the risks associated with developing new technologies for military applications. The challenge for the EMPF is to faciliate the development of advanced power devices that will be applicable to power systems used in future surface ship platforms.

In an effort to overcome the challenges of efficiently and economically cooling the increasing power dissipation densities, the EMPF has teamed with two thermal management companies to develop advanced cooling strategies for the DDG 1000 Power Electronic Modules (PEMs). The EMPF is providing expertise in power electronics packaging design and manufacturing in addition to its knowledge base of Navy applications. Additionally, the team is working with IGBT manufacturers to integrate the advanced thermal management schemes into their packages while maintaining a standard footprint. In this article, the two thermal management strategies will be reviewed and compared to the conventional methods.

Currently, most power modules are designed to be cooled by attaching the module to an external heatsink or cold plate and cooled by forced air or circulated liquid, respectively. As shown in Figure 5-1, the critical layers in the thermal path of  a conventional IGBT module are the IGBT die, the die attach solder, a direct bonded copper (DBC) ceramic substrate, substrate attach solder, metal or composite baseplate, thermal interface material (TIM), and the external coldplate. The ceramic substrate provides electrical isolation between the die and the module baseplate. Aluminum nitride (AlN) is preferred over alumina (Al2O3) because of its higher thermal conductivity. Each layer contributes to the thermal resistance between the die and the ambient through the relationship:

Rt = ρ × t/A

where Rt is the thermal resistance, ρ is the material resistivity, t is the material thickness and A is the contact area. In very general terms, the total thermal resistance is a sum of the resistances of each layer. With the many layers in the heat flow path, this configuration is not capable of adequately cooling devices with power dissipation densities beyond 250 - 300 W/cm².

Recently, two complementary strategies for improving the effectiveness of power module cooling have emerged. One method focuses on reducing the thermal resistance by eliminating layers between the die and the cooling medium and also reducing the thickness and/or thermal resistance of the remaining layers. The second method has been aimed at increasing the efficiency of the cold plate by improving the heat transfer from the body of the heatsink to the coolant. Cross sections of two examples of integrated heatsink concepts that utilize the combination of these methods are shown in Figures 5-2 and 5-3.

Figure 5-2 is a schematic concept of a Normal Flow Cold Plate  (NCP) developed by Mikros Technologies, Inc., Claremont, NH, as integrated within an IGBT package. The thermal path to the cooling fluid is reduced, replacing the baseplate and thermal interface material. The design of the NCP will address a potential mismatch of the Coefficient of Thermal Expansion (CTE) with the ceramic substrate. Another option that would not require CTE matching would be to micromachine the channels into the DBC copper, eliminating the solder interface to the cold plate. The latter option would further reduce the thermal path, as well as increase the efficiency of heat transfer to the cooling fluid.  CTE matching would be less critical due to the thin, compliant copper layer.

The Mikros approach offers a liquid-cooled heat sink with:

• High heat flux capability - can remove in excess of 1000 Watts/cm2 with an approach temperature difference of only 30°C.
• Very low thermal resistance - as low as 0.03°C/(W/cm2)
• High effectiveness - approaching theoretical limit
• Low pressure drop
• Scalability from millimeter sizes to tens of centimeters for IGBT power devices.

Another approach to integrated thermal management for high power devices has been developed by Advanced Cooling Technologies, Inc. (ACT), Lancaster, PA. The Oscillating Flow Heat Sink that they have developed, as integrated in an IGBT package, is shown in Figure 5-3.

ACT’s oscillating liquid heat transfer technology incorporates a mechanical actuator to generate an oscillatory motion of the liquid. The oscillating liquid absorbs heat with a very high efficiency due to the disrupted liquid-wall boundary layers, which increases the effective thermal transfer from the heat source to the heat sink. The heat transfer performance can be controlled by adjusting the amplitude and frequency of the actuator output.

Fluid oscillations have been investigated over a frequency range of 0 to 20 Hz and amplitudes of 0 to 12”. The oscillating flow heat transfer, with a proper combination of oscillating frequency and stroke, can remove heat fluxes in excess of 1,300 W/cm2. This is equivalent to a thermal conductivity value approaching 250,000 W/m-K. For comparison, a 1/8” OD and 12” long copper/water heat pipe can only handle heat fluxes up to 40 W/cm2. Additionally, the thermal conductivities of copper and diamond materials are on the order of 380 W/m-K and 1,200 W/m-K, respectively.

At the current stage of the project, the effort is to demonstrate the effectiveness of the two technologies. The EMPF has identified candidate IGBT modules that would be compatible with the power electronic modules used in the Naval Integrated Power Systems (IPS) project (Empfasis, Aug. 2006). We are working with the manufacturers to develop strategies for technology evaluation and insertion into their product lines. The applications would extend into emerging commercial applications which could include electric vehicles, as an example.

These projects are a part of the EMPF’s continued activity in the Navy’s ship building affordability initiative. By developing collaborative projects that reduce the acquisition costs of ship board electronics, the EMPF can introduce advanced manufacturing processes, improved electronic devices, materials and system technologies. Expanded use of COTS, open systems and an increased use of electronic functional integration continue to be applied to ship board systems that will result in substantial savings to the Navy.
1Leslie, Scott G. “Cooling Options and Challenges of High Power Semiconductor Modules.” Electronics Cooling 12.3 (2006): 20-26.