Simcenter Flotherm Discrete to Uniform Coverage Source Conversion

Why use Total Coverage Approach for Modeling Non-Uniform Die Power Dissipation?

When modeling thousands of discrete die power sources in a thermal analysis model the computational overhead becomes greater than the predictive accuracy that level of detail provides.  If the discrete sources, defined by start and end locations, are represented on a uniform map of heat input areas the model becomes more efficient and provides results consistent with the discrete approach.  There are always exceptions but is generally true.  Consider the die power distribution below where 5,000 sources are modeled as discrete (left) and on a 50x50 uniform grid (right),  both represent a total power dissipation of 50 Watts. 

The power densities above are quite different but the resulting temperatures are very similar, as shown below. 

A comparison of various uniform resolutions with the discrete approach is shown below.  In many situations a relatively course representation provides accurate temperature prediction.

Python Script Prerequisites

In addition to the Python package, Numpy, an input csv file that is formatted consistent with that required for Simcenter Flotherm Die SmartPart discrete source definition, as shown below.

Python Script Usage

The Python script is executed at a Command Prompt or from a Windows PowerShell and has four required inputs. 

• -i : path to CSV in Simcenter Flotherm Die SmartPart discrete source format
• -r : integer value for total coverage resolution
• -x : die x size, in length units from -i
• -y : die y size, in length units from -i

An example of using the script is shown below:

To download the script you may be required to log in.  If there is any feedback please leave a comment.