Matching Thermal Measurements to Simulations

Let's set the scene for a moment. Suppose you have a system that's running too hot and you need to evaluate cooling solutions, such as a fan or liquid cooling. Any cooling solution can be evaluated in simulation, but how do you define the sources and boundary conditions in the simulation model?

Getting the input data for thermal simulations requires an understanding of what are the main contributors of heat in the system. This means you will need to take some measurements with a prototype so you can define sources and boundary conditions in a simulation model. When testing electronics systems, the main heat contributors are determined through measurement with a few different methods, such as thermocouples or infrared cameras.

Once these main factors are known, it is possible to simulate a system as long as the measurement data is used correctly. To see how this works, will examine which measurements to gather in the design and how they are used in setting up a thermal simulation.

Start With Thermal Measurements

Setting up an accurate thermal simulation starts with taking accurate measurements and defining the situation you want to recreate in a simulation. In a thermal simulation, the end goal is to calculate the temperature distribution for a given set of sources in the system. In addition, we would like to determine the time required for changes to occur once the system is perturbed, such as with air flow. This initial temperature distribution simulation can then be used to evaluate redesigns, such as adding a fan into the system.

Start by Defining Requirements 

The first thing to do before taking temperature measurements is to clearly define the situation you want to compare with measurements. For example, you will need to decide on the following aspects of your test case:

• How much power are important components drawing?

• Are specialty materials like thermal gap pads or heat sinks included?

• Once modifications are made, are the measurements in the system repeatable?

These are all basic tasks when setting up thermal test cases for a product. Two common cases to analyze our standard operating conditions and stressed operating conditions. The former examines the typical scenario where a product is being used, while the latter examines the upper end of the rated operating range where the product could be susceptible to failure.

Where to Take Measurements

Measurements in the system should be gathered in a few key areas in order to determine the temperature distribution inside the product.Temperature measurements need to be taken into possible ways:

• Point measurements with thermocouples

• Point measurements with an infrared thermal sensor

• Complete system measurements with a thermal camera

Thermal measurements with an infrared point sensor are point-and-shoot solutions; they give a temperature reading at a specific point in the system. However, just like thermal cameras, they require the enclosure to be fully open in order to access readings from the PCB. Point measurements with a thermocouple are most desirable because these can be wired into a closed package and attached directly to an IC package for a temperature measurement.

Point measurements with a handheld infrared sensor.

When using a group of thermocouples, you should take measurements directly on all of the hot IC packages as these are the packages that will dominate heating in the system. Direct measurements on IC packages will be used later to determine some materials data needed in the thermal simulation.

A temperature reading of the empty space inside the enclosure should also be taken. This is a measurement of the stagnant air in the design and it will be part of the evaluation of the simulation model. Finally, attach a thermocouple to the enclosure surface or measure the enclosure temperature with a thermal camera.

These direct measurements can be captured with an off-the-shelf DAQ unit, such as the Measurement Computing unit shown below.

DAQ unit for capturing thermocouple measurements.

Once hooked up with thermocouples, the design should be monitored over time to examine the approach to equilibrium. Once the equilibrium temperature is reached, the DAQ will record continuous measurements at each of the measurement points in the system.

Sometimes, getting all of these measurements can be difficult when an enclosure has a low profile and there are no open access points for thermocouples. In those systems, it might make more sense to have the enclosure open and have the PCB confined in a slightly larger box in order to make room for thermocouples. Another option is direct imaging with a thermal camera rather than point measurements with thermocouples as this will give surface temperature measurements directly from the main heat-producing components.

Recreate Measurements in Simulation

Before analyzing changes to a design in a simulation, the existing data needs to be used to recreate your existing measurements in a simulation. Recreating the test case in simulation is a form of validation of your simulation model; the physical test case will first be taken as a reference that is used to benchmark the simulation results.

Thermal simulations require solving the heat equation in a closed system, and this requires knowing the thermal conductivity at various points in the system. Integrated circuit packages do not have this data tabulated in the same way as enclosure materials or PCB materials, but it can be determined from the package thermal resistance value in a datasheet and the direct measurements of the temperature during operation.

A thermal simulation will need a value for the heat source (S) in order to predict the steady-state temperature distribution in the product. The image below shows the inputs needed for an integrated circuit simulation package to calculate the heat flux from the source (S):

Heat flux calculation from package thermal resistance, package dimensions, and temperature measurements.

The value of T(ambient) will be the initial condition for the stagnant air inside the enclosure. The other temperature value (T) can be treated as a static value for purposes of defining a heat source in a simulation. For simplicity, you can set T to be your package temperature measurement at equilibrium.

With the determined value of package thermal conductivity (k) and the static package temperature (T), you can now set up a transient thermal simulation with the following settings:

• Initial conditions: 

  • T(ambient) known, allowed to change

  • Temperature of the enclosure body (can be set to T(ambient))

• Boundary conditions:

  • T known from measurement, set as a static value

  • Temperature outside the enclosure (can be set to room temperature)

Alternatively, you can set up a steady-state thermal simulation by taking T(ambient) as the stagnant air temperature and enclosure temperature you measured when the system was in thermal equilibrium. These simulations are faster and will give you a good validation of the simulation model for your system.

An initial thermal simulation result can be used to evaluate modifications to the system.

Modify the System and Re-simulate

Once the system has been simulated and referenced data is obtained, now proposed modifications can be implemented and the system can be re-examined. this can be done alongside new measurements if possible, although such changes are not always possible in such a system. In any case, simulations are the fastest way to start qualifying changes to a design with the intent of reducing thermal load.

For example, the qualified simulation model you use to reproduce thermocouple measurements can then be modified to include a fan. With CFD-thermal co-simulation, it’s possible to add an airflow source and run a transient simulation to examine the effects of different flow rates on the equilibrium temperature during operation. With this approach, you can quickly identify several possible problems that would be difficult to determine in measurement, such as:

• Hot spots with stagnant air

• Areas with low flow

• How inlet and exhaust placement affect airflow

• How enclosure surface temperature is impacted by the above points

• How great is the temperature decrease as a function of airflow

• Exhaust air temperature

Some of these points can be visualized directly on the board surface as typical thermal simulation, but with streamlines overlaid to illustrate the location and amount of airflow. Other cooling measures such as the use of heat sinks, thermal interface material attachment to an enclosure, or a physically larger enclosure can else be examined in these simulations.

Thermal-aware CFD simulation with streamlines shown.

Although the front-end work of gathering thermal data for use in a simulation is time-consuming, it sets up a design team for faster qualification of design changes related to thermal control and management. Ultimately this reduces prototype spins which cost additional time and money, and the simulation model can be continuously modified as new prototypes are built for testing. Growing design teams can quickly see ROI by adding this capability to their suite of design tools.

Whenever you need to set up thermal-aware CFD simulations for your electronic product, use the complete set of system analysis tools from Cadence. Only Cadence offers a comprehensive set of circuit, IC, and PCB design tools for any application and any level of complexity. Cadence PCB design products also integrate with a multiphysics field solver for thermal analysis, including verification of thermally sensitive chip and package designs.

Subscribe to our newsletter for the latest updates. If you’re looking to learn more about how Cadence has the solution for you, talk to our team of experts.