Test Problems

The following test problems have been developed to validate the model by exercising various components of the code and to illustrate the input structure required for various kinds of scenarios. In brief, we have:

• PEKERIS: A simple (two-layer) Pekeris waveguide.

• TWERSKY: The Pekeris wave guide with surface roughness. Demonstrates that the Twersky scatter works properly.

• SCHOLTE: A two-layer waveguide with an elastic bottom which leads to a Scholte wave. Demonstrates that the elastic half-space condition functions correctly.

• DOUBLE: A double-duct problem demonstrating that gradients are handled properly.

• FLUSED: A three-layer problem involving ocean, sediment and half-space. Demonstrates that multiple layers are treated properly.

• ELSED: A three-layer problem with shear properties in the sediment. Demonstrates that elastic media are handled properly.

• ATTEN: A two-layer problem with volume attenuation. Demonstrates that attenuation is handled properly.

• NORMAL: A problem with several density changes to check out the modal normalization in a severe case.

• ICE: A problem with an elastic ice layer to demonstrate that elastic layers above the water column are handled properly.

For each of these cases, we provide the environmental file along with the print-out from KRAKEN . The CPU times printed were obtained on a 0.5 megaflop workstation.

In all cases, the frequency is chosen as and the transmission loss is computed for a source/receiver depth combination of and respectively. The transmission loss plots show an overlay of KRAKEN (solid line), KRAKENC (dotted line) and SCOOTER (dashed line) results. These results have also been checked against the NRL FSTFLD code which agrees to within (usually less).

• PEKERIS
• TWERSKY
• DOUBLE
• SCHOLTE
• FLUSED
• ELSED
• ATTEN
• NORMAL
• ICE