I'm using gpdc for building theoretical dispersion curves. It works very efficiently. But once in a while I get a message like:
**Warning** mode jumping for mode 0 (middle), reducing step ratio to ...
or
**Warning** mode jumping for mode 0 (end), reducing step ratio to ...
I've read your PhD thesis, and as I understand this means that two modes are too close in the secular function, so an automatic quality control measure jumps to the next two modes. Now, does that means that if I'm computing the dispersion curve for mode 0 and get this error, the results I'm obtaining are actually the dispersion curve for mode 2? and what does exactly the (middle) or (end) imply?
thanks!
Jose Colmenares
mode Jumping for gpdc
-
jcolmenares
- Posts: 1
- Joined: Thu Oct 11, 2007 6:03 pm
- Location: Merida, Venezuela
When you get such warnings, and no error at the end, it means that everything is correct. You don't have to worry about them. These warnings dates from the beginning of the development of this tool and of my PhD.
A warning is issued every time a mode jumping is detected after the computation of a dispersion curve. "middle" or "end" is an indication of how the mode jumping was detected. The conventional internal tuning parameters (steps for searching roots) are set to find the solution as quickly as possible. In most case it is ok, but for some cases, the search must be re-initiated with finer parameters.
If you observe several warnings with the search step reduced by a factor 10 at each, leading to an error... it is then impossible to calculate the dispersion curve, reducing the search step does not help. These cases are encountered only for models with low velocity zones (high stiffness over lower stiffness). To analyze why things cannot be computed you can get a grid of all the values of the sub-determinant for all couples (frequency-velocity) with option -grid in gpdc (see gpdc -h for details).
A warning is issued every time a mode jumping is detected after the computation of a dispersion curve. "middle" or "end" is an indication of how the mode jumping was detected. The conventional internal tuning parameters (steps for searching roots) are set to find the solution as quickly as possible. In most case it is ok, but for some cases, the search must be re-initiated with finer parameters.
If you observe several warnings with the search step reduced by a factor 10 at each, leading to an error... it is then impossible to calculate the dispersion curve, reducing the search step does not help. These cases are encountered only for models with low velocity zones (high stiffness over lower stiffness). To analyze why things cannot be computed you can get a grid of all the values of the sub-determinant for all couples (frequency-velocity) with option -grid in gpdc (see gpdc -h for details).