Thermal management of unmanned ground vehicles (UGVs) is more complex than other electronic equipment because they have to operate in harsh environments such as humid tropical rainforests or sandy deserts where moisture as well as dust and sand can compromise the reliability of the control electronics. Regular open enclosures are certainly not an option; instead they need sealed and ruggedized enclosures to also withstand hard shocks and vibrations.

Thus certain cooling methods are either out of the question or only possible with limitations. For example, in general liquid cooling it is possible to transport heat away from the component and cool it at another location with a larger surface and a better convection into the environment. Unmanned aerial vehicles (UAVs) and unmanned underwater vehicles (UUVs) have the advantage of a good coolant flow of either air or water. UGVs in hot desert environments drive over hot sand with the sun shining on their top side—not the best conditions for any cooling systems.

Efficient cooling methods range from the basic principles of heat transfer to some more costly and more complex physics that can be simulated using computational fluid dynamics (CFD) software. There are three basic cooling mechanisms: conduction, convection, and radiation. Two other methods can be considered too but are more of a hybrid or model more complex physics: advection and phase transition.

Advection is the movement of heat from one point to another, such as heated water run to a heat exchanger, and requires a velocity that is usually provided with the help of a water pump. But it works on the same basic principles, to conduct the heat into the fluid and back out of it.

Phase transition is actually a very efficient method that even our body applies when we get hot, either from the environment or if under high load in a workout at a gym. Our body starts to perspire and the sweat evaporates, and this creates a cooling effect from the energy taken to evaporate the sweat. UGVs can’t sweat, but there are methods applied in various applications where the phase change can be used to cool or control the temperature of a component.

Figure 1. The boiling curve shows qualitatively the dependency of the heat flux on the temperature ΔT on a logarithmic scale. The graph is split in the various regions of the boiling states I‒V and their transition points A‒E.
For example, phase-change material (PCM) is used in modern electric vehicles (EVs) that reduce the critical temperature with the help of this buffering effect. Of course, phase-change materials only work for a certain time and range. We all know that water when melting stays at exactly the freezing temperature and doesn’t get hotter until the ice is molten but, after that, the temperature increases again. This is the same principle for phase-change material, but instead of water, some gels are used that are solid up to a certain temperature where they keep the temperature constant—up to the point where the gel is molten and then increases also in temperature. This technology applies mostly as a peak load buffer.

Another phase change is the boiling of the coolant (Figure 1). This requires a special coolant or mixture to meet a certain boiling range or temperature that suits the desired maximum design temperature of the component. The boiling effect is a sensitive state because as the coolant temperature reaches the transition from single-phase convection to partial nucleate, boiling the coolant will start to form small bubbles that then detach from the surface and rise up. The bubble doesn’t transport the heat, rather it’s the coolant flow that is generated near the wall from the bubble detaching and moving away from the wall. The further the temperature increases, the stronger the boiling gets until it reaches a point where the slope of the increasing heat flux decreases again.

From this point on, we are in the fully developed nucleate boiling range. This ranges up to the maximum heat flux where it then flips and the heat flux decreases again. We would not want to get over that point because suddenly the heat flux decreases again as the temperature increases and that’s not good for the cooling of our device. The zone above the critical heat flux is the transition boiling zone which then enters the film boiling; however that zone is in even higher temperature ranges. This method is used in modern cars’ internal combustion engine water jackets that cool the cylinder block and head.

The third phase-change method that can be used for cooling is evaporation, as we mentioned already. Now, I said that evaporation is something our UGVs cannot use as many living creatures do, but then, humans are creative. We find ways to use this effect even for machines. The application of spray cooling is exactly what most resembles the sweating of a human. In spray cooling, the coolant is sprayed with a nozzle onto the hot surface that wets the surface which is then evaporated and cooled down, until it changes phase back to liquid (again elsewhere in the cooling loop). This cooling cycle is similar to regular liquid cooling but with a phase change. This type of phase change is already applied in some electronics cooling applications. In some ways, it is similar to a heat pipe, where the coolant evaporates at the hot end and condenses at the cold end. As the coolant either flows back because of gravity or when a wick is used, the coolant is sucked back as a result of the capillary effect.

So besides the basic principles of fans and conduction and natural convection, higher more complex cooling methods find increased interest in applications that were not used even several years ago. Advanced cooling is needed with the increased heat generated by complex military designs where cooling is often not that simple, especially when harsh environments prohibit certain mechanisms, making them operate less effectively.

Case Study

Figure 2. An example of an Azonix rugged embedded computer.
The following example illustrates the types of challenges faced when designing electronic equipment for the types of environments that UGVs operate in. Engineers at Azonix, a division of Crane Co., used the Mentor Graphics® FloEFD® computational fluid dynamics (CFD) thermal simulation software when designing the Terra embedded computer. The Terra computer is designed to be completely sealed from the extreme (or harsh) elements and for use in very hot environments (Figure 2). The simulations enabled them to reduce the number of thermal prototypes they had to make from 12 to 1.

The engineers used their CAD geometry with the CFD software and defined the heat dissipation sources, material properties, and the ambient temperature outside the enclosure at the product’s design limit of 60 °C. They then defined the goals and performed thermal simulation. The CFD software analyzed the CAD model, automatically identified fluid and solid regions, and defined the entire flow space without interaction and without adding extra objects to the CAD model. The software generated simulation results in roughly five minutes. The results revealed that temperatures on the surfaces of key components exceeded the allowable limit of 90 °C.

The conduction paths from the heat dissipating components to the heatsink and heatsink geometry were the primary design parameters that offered an opportunity to improve thermal performance. The cross-section of the heat spreader was increased and changed from aluminum to copper. Gap-type thermal interface material was inserted at the interfaces between the components and the heat spreader. The thermal interface material was modeled as a contact resistance, reducing the number of cells, rather than conduction through material.

Figure 3. The CFD thermal simulation shows the air temperature distribution (left), and the velocity magnitude and field in the Azonix model (right).
These changes substantially reduced the surface temperatures on the dissipating components, though still not enough to meet the thermal requirements. They then optimized the design of the heatsink. After roughly six iterations, in each case changing the spacing and height of the fins, the heatsink was optimized and the internal component temperatures held to a minimum.

This is just one example of engineers overcoming, in little time and using sophisticated simulation tools, the design challenges in military and aerospace applications.

This article was written by Boris Marovic, Industry Manager for Aerospace and Defense, Mentor Graphics Mechanical Analysis Division (Frankfurt, Germany). For more information, Click Here .