Filter feeding in devil rays is highly sensitive to morphology
Data files
Dec 28, 2024 version files 1.61 GB
-
kahane_rapport_etal_2024.tar.gz
1.61 GB
-
README.md
2.18 KB
Abstract
Mobulid rays (manta and devil rays) use a highly-specialized filtering apparatus to separate plankton food particles from large volumes of seawater. Recent studies have indicated that captive vortices form within the microscale pores of the filter, which enhance filtration efficiency through a novel mechanism referred to as ricochet separation. The high throughput and clog resistance of this filtration process has garnered the attention of the engineering community and inspired the development of several engineered filtration systems. However, it is still unclear how changes to the filter morphology influence the surrounding flow patterns and filtration efficiency. We address this question by examining the flow fields around and filtering properties of mobulid filters with systematically varied morphologies, using a combination of computational fluid dynamics and experiments on physical models. Our results indicate that the morphological features critical to the performance of the mobulid filtering structure are substantially different from those expected for a conventional sieve filter. While the pore size is the principal determinant of the size-dependent filtration efficiency in a sieve filter, we found that the captive vortices in a mobulid filter grow or shrink to fill the pore, and changes in the pore size have relatively little effect on filtration efficiency. In contrast, the filtration efficiency appears to be highly sensitive to the orientation of the filter lobes (microscale plate-like structures). These results provide a foundation for interpreting the morphological differences between species and also for generating optimized bioinspired designs.
https://doi.org/10.5061/dryad.280gb5n0s
Description of the data and file structure
Computational fluid dynamics (CFD) simulations and particle image velocimetry (PIV) experiments examining the biomechanics software filter feeding in devil rays.
Files and variables
File: kahane_rapport_etal_2024.tar.gz
Description: Scripts and model files for CFD and PIV, organized as
- cfd_models : input files for CFD simulations
- freestream_vel : simulations varying freestream velocity
- lobe_angle_offset : simulations varying lobe angle
- lobe_period : simulations varying lobe period
- lobe_tip_angle_offset : simulations varying lobe tip angle
- lobe_tip_length : simulations varying lobe tip length
- lobe_tip_radius : simulations varying lobe tip radius
- cfd_scripts : MATLAB scripts for generating CFD input files
- analyze_trace_rev.m : analyze time-reversed fluid trajectories
- calc_percent_filtered.m : calculate percentage of particles 'filtered'
- calc_solid_trace_results.m : calculate particle filtration for simulation series
- calc_trace_rev_results.m : calculate time-reversed fluid trajectories for simulation series
- generate_model_initial.m : write CFD input file for initial simulation
- generate_model_restart.m : write CFD input file for restart simulation
- generate_model_spoiler_basic_config : generate model configuration for spoiler-like lobes
- generate_model_wing_basic_config : generate model configuration for wing-like lobes
- generate_run_trace_batch_script.m : generate batch processing script for particle tracing
- generate_trace_params_solids.m : generate parameters file for particle tracing
- particle_sim : source code for particle tracing simulations
- build : build folder for code
- c_source : source code for library
- Makefile : build script for library
- zargs : command-line processing library
- piv_scripts : scripts for DPIV analysis
- osiv_corr_params.txt : parameters file for osiv_corr
- processRawResults.m : MATLAB file for processing DPIV results
The deposited material provides the scripts and models used for computational fluid dynamics (CFD) simulations and particle image velocimetry (PIV) experiments. Further details may be found in the associated research article.
