-
-
July 27, 2023 at 5:03 pm
tlevaur
SubscriberHello,
I am working on a simulation to optimize the transmission through an AR-coated III-V waveguide facet. I have an optimization working on three variables, the angle of the interface, the refractive index of the AR coating, and the thickness of the coating. I will run the optimization with a generation size of 20 and about 1-3 simulations of the generation will fail unexpectedly. I have gone to the offending simulations (Number 13 in this case) and attempted to run it manually and received the exact same error, so I do not believe it has anything to do with licensing (as I saw suggested in another post somewhat related to this one).
Originally, I had the mesh accuracy set to 5. Setting it to 6 allows the simulation to run, so I tried to run the optimization again with the higher accuracy and still had 1-3 simulations fail. Also, this is not a desirable workaround as it slows down each generation.Â
I know there are likely smarter ways of designing said single-layer AR coating, but I see no reason why my approach does not work for certain combinations of the three parameters.
Any help would be greatly appreciated.
Edit: posted script did not appear as it should've
-
July 27, 2023 at 5:19 pm
-
July 31, 2023 at 10:23 pm
Guilin Sun
Ansys EmployeeI thought you have resolved the issue. Sorry to reply late.
Error code 5 might be due to temporary server/license access issue. THis could happen when the network adaptor fails to work for some time due to uknown reasons;Â
To isolate the cause, please change the optimization to sweep. If the sweep has the same error, then we can identify the file, and its settings as well as result matrix if any. You may try at least 5 simulation files at each generation.
Â
-
August 1, 2023 at 8:56 pm
tlevaur
SubscriberI did not resolve this issue. That script I posted contains the relevant simulation parameters that produce the invalid matrix dimensions error.Â
-
August 2, 2023 at 5:36 pm
Guilin Sun
Ansys EmployeeWe are not allowed to download files in this forum. But I had a look. I did not find anything significantly wrong. However since you did not mention how do you set the angles, it is hard to comment.
Current you use polygons. You will need to have the angle in model or in structure group to sweep/optimize it.
You may also try to use mesh accuracy 2, but with override mesh for the max title angle region, and try again.
Error code 5Â is a system error, probably from c++ library. You can search online for that error message.
-
- The topic ‘Invalid Matrix Error’ is closed to new replies.
- Difference between answers in version 2024 and 2017 lumerical mode solution
- Import material .txt file with script
- Absorption cross-section of AuNR excited by prism-based TIR
- How to measure transmission coefficients on a given plane .
- TE Polarization Fraction + Waveguide TE/TM Fraction
- Edge Coupler EME Example Issue
- Lumerical Python API Topology Optimization Error 2025 R1
- 2D inverse design grating coupler
- FDE : Simulation waveguides with electrodes
- Loss of lithium niobate phase modulator
-
2407
-
930
-
599
-
591
-
564
© 2025 Copyright ANSYS, Inc. All rights reserved.