Parameter sweep in transient heat simulations

Dear Lumerical Support,

I am running transient heat simulations, and I am trying to plot the temperature in the nanostructure as a function of peak fluence of the laser light illuminating the nanostructure. To this end, I run a parameter sweep and change scaling factor in my heat simulation. After I run the sweep, the executed files are saved in a separate folder. Then, to extract the temperature as a function of the scaling factor, I load files back into the parameter sweep and try to extract the temperature by using the script entitled “getsweepdata”.
However, when I try to run the script “getsweepdata”, I receive the following error message: “The optimization or parameter sweep object ‘Pin’ has no results.”

Interestingly, when I perform the same procedure for steady state simulations, everything works fine. Could you please help me with identifying the problem.
The relevant files are the optical simulation file “2D_Au_ITO_Al2O3_optical”, the script to extract the absorptions densit from this optical simulation file “extracting_absorption_density_1500”, heat simulation file “2D_Au_ITO_Al2O3_optical_heat”, two files saved from the parameter sweep “Pin_1” an “Pin_2” as well as the script to extract the data from the sweep “getsweepdata”. The files can be downloaded from here:


Dear Ruzan,

Sorry for the long wait. Here, the issue is due to the results obtained for different input powers, having different sizes:

In your file, you can notice there is no red icon on the sweeping object after it is run. This indicates that there is no data stored into the object (which explains the error message you mentioned). To investigate this, the warning message prompted in the parameter sweep status window is insightful:

This message indicates that the data is not being saved due to the results of different simulations having different sizes (as mentioned in the beginning). The issue is similar to what has been discussed in this post:

In your simulation, the problem specifically stems from the number of time steps being different in the two transient simulations, which in turn leads to the results in temperature monitors having different sizes:

This should explain why running a steady-state simulation doesn’t cause any issues. Depending on the results you are interested in, you might decide to take different approaches to address this, such as settling for a steady-state simulation, extracting the data manually or by using scripting, etc.

Kind regards,