Hi, I use a single lens to focus the light to a point. Some heat sources were loaded on the single lens. I used WorkBench to simulate the heat distribution and exported the heat distribution, and then imported the finite element thermal data into Zemax STAR. The ambient temperature is 22°C. After loading the heat source, the lowest temperature of the lens surface will not be lower than 22°C, which is normal. However, we found that by checking the gradient refractive index fitted by STAR (database - optical materials - GRIN profile), we found that the temperature corresponding to the fitted refractive index was lower than 22°C, which is obviously wrong! Is there any way to correct this?
Attached is the test .Zar file. Thanks!
Page 1 / 1
Hi @Haokun.Ye ,
thank you for your interest in STAR. First, I would recommend defining a second wavelength before fitting the data. This ensures that a temperature fit is generated, which simplifies checking the fit with the STAR System Viewer.
If you switch to the Voxel Draw Mode, you can see based on the color bar that there are areas with a temperature slightly below 22°C in the fit. This is most likely caused by the type of mesh that you are using. The fitting algorithm works best with unstructured meshes (tetrahedrons) with a uniform element size so that the number of nodes on the surface is somewhere around ten-thousand. The mesh you are using is quite coarse and the distance between node varies extremely, which both are factors that cause the fitting algorithm to struggle.
I hope that this answers your question,
Best Regards
Hi @Matthias.Schlich , thank you very much for your reply, I think I can understand that the grid is causing the refractive index fitting problem. However, my optical system is a single-wavelength system. Is there any special purpose of defining the second wavelength? Thanks!
Adding the second wavelength is for testing purposes only, having multiple wavelengths triggers a different fitting method. Instead of converting from temperature to index before fitting (what happens for a single wavelength), the temperature is fitted and the conversion to index happens “on the fly” during raytracing. This allows you to directly inspect the fitted temperature and compare it with the original FEA data.
@Matthias.Schlich Ok I understand, thank you very much for your help!
Since this is the first time that I’ve heard that adding a second wavelength can increase the accuracy of the fit I have some questions:
Can this second wavelength be very close to the primary? As an example let’s say the primary is 632.8nm, will the quality of the fit be better if we add a second wavelength of 632.81nm?
Once the fit is created, using this second wavelength, can the second wavelength be turned off for any analyses, with no effect on the fit? Based on your description it looks like the answer to this is no because it is converting the index ‘on the fly’ as you say do this second wavelength always needs to be present. Is that right?
@KnottMyName The fit accuracy will not change based on the number of wavelengths. The reason why I suggested to add a wavelength before the fit is solely to be able to compare the fitted temperature with the original. The two fit methods are described in the help file at The STAR Tab » FEA Fitting Process and RBMs.
Regarding your questions, to ensure that the temperature is fitted, the actual value of the second wavelength does not matter, only that there is more than one.
Turning the second wavelength off once the temperature is fitted is actually a great idea. The 2nd wavelength is only needed to trigger the different fit method (when you load the FEA data). Afterwards, it can be turned off again so that it does not interfere with your analyses.
