Code, data, and videos from: Gyroscopic spiders: Web mechanical intelligence facilitates prey sensing
Data files
Jul 16, 2026 version files 93.81 MB
-
Finite_Element_Abaqus_Files.zip
65.35 MB
-
README.md
4.37 KB
-
SV1.mp4
28.16 MB
-
SV2.mp4
304.62 KB
Abstract
The code, data, and video files are provided as supporting materials for the manuscript titled: "Web Mechanical Intelligence Facilitates Prey Sensing." The example files are generated during the study of the robustness of the pitch-roll sensing mechanism, as observed in videos of prey striking natural webs with spiders, and verified through more than 18,000 computational simulations and 74 experiments with artificial webs. Details are provided in the main manuscript. Three distinct spider webs are used, with varying degrees of narrowness and eccentricity. All webs shared the same structural links, consisting of 56 radial threads, 19 spiral turns, and a secondary frame that maintains mechanical integrity. The full web geometries were generated using the Force Density Method to satisfy static equilibrium and to follow tension gradients and structural layouts reported in biological studies. The force density method linearizes the static force equilibrium equations, enabling the fast form-finding of the web’s network with desired narrow (Nar) and eccentric (Ecc) shape parameters using tension gradients. Python code was developed to generate webs with desired shapes using the force-density method. The code also imports the geometries to Abaqus, a commercial Finite Element Method (FEM) package. Details are provided in the Materials and Methods section of the manuscript. The video SV1 is an example of an unfortunate (Gryllodes sigillatus) cricket hitting an orb web (Argiope trifasciata), where a spider gyroscopic motion is observed followed by reorintation toward the cricket. The video is analyzed using the digital image correlation method, as detailed in the Supplementary Materials. Video SV2 shows the similar dynamics in an artificial web with an equivlent spider mass at the hub and simulated prey impacts using miro-hammer sensor.
Dataset DOI: 10.5061/dryad.547d7wmpp
Description of the data and file structure
SV1.mp4
SV2.mp4
Finite_Element_Abaqus_Files.zip
Access information
Movie SV1--- Live Spider Pitch–Roll Response To Impact
A Gryllodes sigillatus cricket impacts the orb web of Argiope trifasciata. The digital image correlation (DIC) analysis used SV1.mp4 to track the spider’s body motion. Directional cue strength shows the spider pitches toward the prey, with pitch 14.8× greater than roll.
Movie SV2--- Artificial Web Dynamics Dominate Vibration Deflections Due To Impact
An artificial web with representative spider and prey masses is excited using a micro‑hammer. High‑speed imaging and DIC time‑history data yield frequency response functions. Results summarize how prey mass alters low‑frequency modes and directional cues.
Finite_Element_Abaqus_Files--- Artificial Web Dynamics simulations and data
Input files (.inp) starting with "Web_e" are base input files compatible with Abaqus that get modified and executed using python scripts. Before these files run you will have to update the file roots in the scripts. The .inp file can also be run directly from Abaqus CAE, in which case the model yields the mode shapes of a web without prey.
The input file contains the following:
- *Node. all nodes describing the web.
- *NSET, are sets of nodes. NSET=MASS_LOC_ALL: all spiral nodes that intersect with a radial. The Python script "Run_model_ecc#_2.py" evaluates the modes shapes and natural frequency while iteratively moving the mass element through this node set.
- *ELEMENT. All elements making up the web. The elements link node numbers. Divided into three groups: SPIRAL_ELEMENTS, RADIAL_ELEMENTS, MOORING_ELEMENTS.
- *MATERIAL. Here linear material properties are described to spiral, radial, and mooring elements. See paper for further information. Compression is disabled, as the web sections will act like strings.
- *SOLID SECTION. Respective cross sectional area is described to spiral, radial, and mooring elements.
- *BOUNDARY. The Connection points are fixed
- *INITIAL CONDITIONS. Here pre-stress is applied to all elements such that the web is in equilibirum as the web is imported. The force densities of elements were tuned such that both web shape and web tension are realistic.
- *STEP. Here a static step prepares the frequency step and checks equilibirum. Then the Frequency step extracts natural frequencies and mode shapes.
The scripts "Run_model_ecc#_2.py" were used to generate the data for figures 3c, S2, S3, S4, S5, S6 and S7 in the related article Masmeijer et al. (2026).
The nested loop evaluates the input file in Abaqus subsequently at all nodes in nodeset MASS_LOC_ALL, for the masses in in the list m_prey_vec. The function creates a set of nodes at the spider's leg positions and at the spider's mass center. The resulting natural frequencies are saved in a .csv file. The modes are only saved at the leg positions to save space. Full output models are saved for the node numbers as keys in the Full_save_dict dictionary.
Run Python script in an Abaqus environment from command line: "abaqus python Run_model_ecc#_2.py"
Script inputs: None
Script outputs: .csv files with natural frequencies and leg displacements, and .odb files for selected mass locations
"Run_batch_models_masses.py" was used to generate data for figure 3a and 3b in the related article Masmeijer et al. (2026)
Run Python script in an Abaqus environment from command line: "abaqus python Run_batch_models_masses.py [ecc]"
Script inputs: [ecc] (either 0, 1, or 2)
Script outputs: .csv files with natural frequencies and leg displacements. .odb files for selected mass locations
One example output database file is included (Web_e1_g0_maxe5_mine15_ss1_m3.50782182348e-10_15.odb), which can be viewed using Abaqus.
References:
T.C.P. Masmeijer, J. Slavic, P. De , T.A. Blackledge, E. Habtour, Gyroscopic Spiders: Web Mechanical Intelligence Facilitates Prey Sensing, Scientific Reports, (in review).
T. Masmeijer, C. Swain, J. Hill, E. Habtour, Programming tension in 3D printed networks inspired by spiderwebs. Materials & Design p. 115281 (2026).