Raw data of scanning electron microscopy (SEM), atomic force microscopy (AFM), force spectroscopy, data analysis and plotting, optical microscopy, and finite element simulations (FEA) for our manuscript. File Formats AFM raw data is provided in Gwyddion format, which can be viewed using the Gwyddion AFM viewer, which has been released under the GNU public software licence GPLv3 and can be downloaded free of charge at http://gwyddion.net/ Optical microscopy data is provided in JPEG format SEM raw data is provided in TIFF format Data analysis codes were written in MATLAB (https://www.mathworks.com/products/matlab) and stored as *.m files Imported raw data to MATLAB and saved MATLAB data were stored as MATLAB multidimensional arrays (MATLAB “struct” data format, *.mat files) FEA results were saved as text files, .txt files) Data (Folder Structure) The data in the dataverse is best viewed in Tree mode. Read me file.docx More Explanations of analysis in docx format. Figure 1 Figure 1 - panel b.jpg (5.5 MB) Optical micrograph (JPEG format) Figure 1 - panel c - AFM Raw Data.gwy (8.0 MB) AFM raw data (Gwyddion format) Figure 1 - panel e - P0_Force-curve_raw_data.txt (3 KB) Raw force-displacement data at P0 (text format) Figure 1 - panel e - Px_Force-curve_raw_data.txt (3 KB) Raw force-displacement data at Px (text format) Figure 1 - panel e - Py_Force-curve_raw_data.txt (3 KB) Raw force-displacement data at Py (text format) Figure 1 - panel e - P0_simulation_raw_data.txt (12 KB) FEA simulated force-distance data at P0 (text format) Figure 1 - panel e - Px_simulation_raw_data.txt (12 KB) FEA simulated force-distance data at Px (text format) Figure 1 - panel e - Py_simulation_raw_data.txt (12 KB) FEA simulated force-distance data at Py (text format) Figure 1 - panel e - FCfindc.m (2 KB) MATLAB code to calculate inverse optical lever sensitivity (InverseOLS) of AFM cantelever (matlab .m format) Figure 1 - panel e - FreqFindANoise_new.m (2 KB) MATLAB code to calculate white noise constant, A (Explained in the Read me file) (matlab .m format) Figure 1 - panel e - FreqFindQ_new.m (4 KB) MATLAB code to calculate Q factor of the AFM cantelever (matlab .m format) Figure 1 - panel e - FCkeff.m (2 KB) MATLAB code to calculate the effective spring constant k of the AFM cantelever (matlab .m format) Figure 1 - panel e - FCimport.m (7 KB) MATLAB code to import raw force-displacement data into MATLAB (matlab .m format) Figure 1 - panel e - FCForceDist.m (2 KB) MATLAB code to convert raw force-displacement data into force-distance data (matlab .m format) Figure 1 - panel e - Figure 1- Panel e - data.mat (6 KB) MATALB struct data file for calibrated force-distance data at all indentation points (matlab .mat format) Figure 1 - panel e - Panel_e_MatlabCode.m (6 KB) MATALB code for plotting experimental and simulated force curves in panel e (matlab .m format) Figure 1 - panel e - Read me file - force curve calibration.docx (14 KB) Explains force curve calibration (.docx format) Figure 1 - panel e - Read me file - lever spring constant calibration.docx (14 KB) Explains AFM lever spring constant calibration (.docx format) Figure 2 Figure 2 - panel a - MATLAB data.mat (2.6 KB) MATALB data file for simulated data (matlab .mat format) Figure 2 - panel b - MATLAB data.mat (2.4 KB) MATALB data file for simulated data (matlab .mat format) Figure 2 - panel a - simulation raw data.txt (5.0 KB) Raw simulation data: xyz coordinates of the nodes of deformed FEA mesh (text format) Figure 2 - panel b - simulation raw data.txt (5.0 KB) Raw simulation data: xyz coordinates of the nodes of deformed FEA mesh (text format) Figure 2 - panel ab - MATLABcode.m (1.0 KB) MATALB code for plotting panel a b figures (matlab .m format) Figure 2 - panel c - Degree of Anisotropy datacode.m (1.0 KB) MATALB code for plotting panel c graph (matlab .m format) Figure 3 Figure 3 - panel a - App_curve_1_raw_data.txt (35 KB) Raw force-displacement data approach curve 1 (text format) Figure 3 - panel a - App_curve_2_raw_data.txt (34 KB) Raw force-displacement data approach curve 2 (text format) Figure 3 - panel a - App_curve_3_raw_data.txt (34 KB) Raw force-displacement data approach curve 3 (text format) Figure 3 - panel a - App_curve_4_raw_data.txt (34 KB) Raw force-displacement data approach curve 4 (text format) Figure 3 - panel a - Ret_curve_1_raw_data.txt (35 KB) Raw force-displacement data of retract curve 1 (text format) Figure 3 - panel a - Ret_curve_2_raw_data.txt (35 KB) Raw force-displacement data of retract curve 2 (text format) Figure 3 - panel a - Ret_curve_3_raw_data.txt (35 KB) Raw force-displacement data of retract curve 3 (text format) Figure 3 - panel a - Simulation_raw_data-part 1.txt (43 KB) simulated force-displacement data of -part 1 (text format) Figure 3 - panel a - Simulation_raw_data-part 2.txt (43 KB) simulated force-displacement data of -part 2 (text format) Figure 3 - panel a - FCfindc.m (2 KB) MATLAB code to calculate inverse optical lever sensitivity (InverseOLS) of AFM cantelever (matlab .m format) Figure 3 - panel a - FreqFindANoise_new.m (2 KB) MATLAB code to calculate white noise constant, A (Explained in the Read me file) (matlab .m format) Figure 3 - panel a - FreqFindQ_new.m (4 KB) MATLAB code to calculate Q factor of the AFM cantelever (matlab .m format) Figure 3 - panel a - FCkeff.m (2 KB) MATLAB code to calculate the effective spring constant k of the AFM cantelever (matlab .m format) Figure 3 - panel a - FCimport.m (7 KB) MATLAB code to import raw force-displacement data into MATLAB (matlab .m format) Figure 3 - panel a - FCForceDist.m (2 KB) MATLAB code to convert raw force-displacement data into force-distance data (matlab .m format) Figure 1 - panel e - Read me file - force curve calibration.docx (14 KB) Explains force curve calibration (.docx format) Figure 1 - panel e - Read me file - lever spring constant calibration.docx (14 KB) Explains AFM lever spring constant calibration (.docx format) Figure 3 - panel b - SEM Raw Data.tiff (9 KB) SEM raw image of broken silk membrane due to extreme indentation (.tiff format)
more »
« less
This content will become publicly available on October 1, 2024
AI-dente: an open machine learning based tool to interpret nano-indentation data of soft tissues and materials
Nano-indentation is a promising method to identify the constitutive parameters of soft materials, including soft tissues. Especially when materials are very small and heterogeneous, nano-indentation allows mechanical interrogation where traditional methods may fail. However, because nano-indentation does not yield a homogeneous deformation field, interpreting the resulting load–displacement curves is non-trivial and most investigators resort to simplified approaches based on the Hertzian solution. Unfortunately, for small samples and large indentation depths, these solutions are inaccurate. We set out to use machine learning to provide an alternative strategy. We first used the finite element method to create a large synthetic data set. We then used these data to train neural networks to inversely identify material parameters from load–displacement curves. To this end, we took two different approaches. First, we learned the indentation forward problem, which we then applied within an iterative framework to identify material parameters. Second, we learned the inverse problem of directly identifying material parameters. We show that both approaches are effective at identifying the parameters of the neo-Hookean and Gent models. Specifically, when applied to synthetic data, our approaches are accurate even for small sample sizes and at deep indentation. Additionally, our approaches are fast, especially compared to the inverse finite element approach. Finally, our approaches worked on unseen experimental data from thin mouse brain samples. Here, our approaches proved robust to experimental noise across over 1000 samples. By providing open access to our data and code, we hope to support others that conduct nano-indentation on soft materials.
more »
« less
- NSF-PAR ID:
- 10451286
- Date Published:
- Journal Name:
- Soft Matter
- ISSN:
- 1744-683X
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Hydrogels are a class of soft, highly deformable materials formed by swelling a network of polymer chains in water. With mechanical properties that mimic biological materials, hydrogels are often proposed for load bearing biomedical or other applications in which their deformation and failure properties will be important. To study the failure of such materials a means for the measurement of deformation fields beyond simple uniaxial tension tests is required. As a non-contact, full-field deformation measurement method, Digital Image Correlation (DIC) is a good candidate for such studies. The application of DIC to hydrogels is studied here with the goal of establishing the accuracy of DIC when applied to hydrogels in the presence of large strains and large strain gradients. Experimental details such as how to form a durable speckle pattern on a material that is 90% water are discussed. DIC is used to measure the strain field in tension loaded samples containing a central hole, a circular edge notch and a sharp crack. Using a nonlinear, large deformation constitutive model, these experiments are modeled using the finite element method (FEM). Excellent agreement between FEM and DIC results for all three geometries shows that the DIC measurements are accurate up to strains of over 10, even in the presence of very high strain gradients near a crack tip. The method is then applied to verify a theoretical prediction that the deformation field in a cracked sample under relaxation loading, i.e. constant applied boundary displacement, is stationary in time even as the stress relaxes by a factor of three.more » « less
-
A comparative study is presented to solve the inverse problem in elasticity for the shear modulus (stiffness) distribution utilizing two constitutive equations: (1) linear elasticity assuming small strain theory, and (2) finite elasticity with a hyperelastic neo-Hookean material model. Assuming that a material undergoes large deformations and material nonlinearity is assumed negligible, the inverse solution using (2) is anticipated to yield better results than (1). Given the fact that solving a linear elastic model is significantly faster than a nonlinear model and more robust numerically, we posed the following question: How accurately could we map the shear modulus distribution with a linear elastic model using small strain theory for a specimen undergoing large deformations? To this end, experimental displacement data of a silicone composite sample containing two stiff inclusions of different sizes under uniaxial displacement controlled extension were acquired using a digital image correlation system. The silicone based composite was modeled both as a linear elastic solid under infinitesimal strains and as a neo-Hookean hyperelastic solid that takes into account geometrically nonlinear finite deformations. We observed that the mapped shear modulus contrast, determined by solving an inverse problem, between inclusion and background was higher for the linear elastic model as compared to that of the hyperelastic one. A similar trend was observed for simulated experiments, where synthetically computed displacement data were produced and the inverse problem solved using both, the linear elastic model and the neo-Hookean material model. In addition, it was observed that the inverse problem solution was inclusion size-sensitive. Consequently, an 1-D model was introduced to broaden our understanding of this issue. This 1-D analysis revealed that by using a linear elastic approach, the overestimation of the shear modulus contrast between inclusion and background increases with the increase of external loads and target shear modulus contrast. Finally, this investigation provides valuable information on the validity of the assumption for utilizing linear elasticity in solving inverse problems for the spatial distribution of shear modulus associated with soft solids undergoing large deformations. Thus, this work could be of importance to characterize mechanical property variations of polymer based materials such as rubbers or in elasticity imaging of tissues for pathology.more » « less
-
Abstract We present a data-driven workflow to biological tissue modeling, which aims to predict the displacement field based on digital image correlation (DIC) measurements under unseen loading scenarios, without postulating a specific constitutive model form nor possessing knowledge of the material microstructure. To this end, a material database is constructed from the DIC displacement tracking measurements of multiple biaxial stretching protocols on a porcine tricuspid valve anterior leaflet, with which we build a neural operator learning model. The material response is modeled as a solution operator from the loading to the resultant displacement field, with the material microstructure properties learned implicitly from the data and naturally embedded in the network parameters. Using various combinations of loading protocols, we compare the predictivity of this framework with finite element analysis based on three conventional constitutive models. From in-distribution tests, the predictivity of our approach presents good generalizability to different loading conditions and outperforms the conventional constitutive modeling at approximately one order of magnitude. When tested on out-of-distribution loading ratios, the neural operator learning approach becomes less effective. To improve the generalizability of our framework, we propose a physics-guided neural operator learning model via imposing partial physics knowledge. This method is shown to improve the model's extrapolative performance in the small-deformation regime. Our results demonstrate that with sufficient data coverage and/or guidance from partial physics constraints, the data-driven approach can be a more effective method for modeling biological materials than the traditional constitutive modeling.more » « less
-
SUMMARY Anisotropy of remanent magnetization and magnetic susceptibility are highly sensitive and important indicators of geological processes which are largely controlled by mineralogical parameters of the ferrimagnetic fraction in rocks. To provide new physical insight into the complex interaction between magnetization structure, shape, and crystallographic relations, we here analyse ‘slice-and-view’ focused-ion-beam (FIB) nano-tomography data with micromagnetic modelling and single crystal hysteresis measurements. The data sets consist of 68 magnetite inclusions in orthopyroxene (Mg60) and 234 magnetite inclusions in plagioclase (An63) were obtained on mineral separates from the Rustenburg Layered Suite of the Bushveld Intrusive Complex, South Africa. Electron backscatter diffraction was used to determine the orientation of the magnetite inclusions relative to the crystallographic directions of their silicate hosts. Hysteresis loops were calculated using the finite-element micromagnetics code MERRILL for each particle in 20 equidistributed field directions and compared with corresponding hysteresis loops measured using a vibrating sample magnetometer (VSM) on silicate mineral separates from the same samples. In plagioclase the ratio of remanent magnetization to saturation magnetization (Mrs/Ms) for both model and measurement agree within 1.0 per cent, whereas the coercivity (Hc) of the average modelled curve is 20 mT lower than the measured value of 60 mT indicating the presence of additional sources of high coercivity in the bulk sample. The VSM hysteresis measurements of the orthopyroxene were dominated by multidomain (MD) magnetite, whereas the FIB location was chosen to avoid MD particles and thus contains only particles with diameters <500 nm that are considered to be the most important carriers of palaeomagnetic remanence. To correct for this sampling bias, measured MD hysteresis loops from synthetic and natural magnetites were combined with the average hysteresis loop from the MERRILL models of the FIB region. The result shows that while the modelled small-particle fraction only explains 6 per cent of the best fit to the measured VSM hysteresis loop, it contributes 28 per cent of the remanent magnetization. The modelled direction of maximal Mrs/Ms in plagioclase is subparallel to [001]plag, whereas Hc does not show a strong orientation dependence. The easy axis of magnetic remanence is in the direction of the magnetite population normal to (150)plag and the maximum calculated susceptibility (χ*) is parallel to [010]plag. For orthopyroxene, the maximum Mrs/Ms, maximum χ* and the easy axis of remanence is strongly correlated to the elongation axes of magnetite in the [001]opx direction. The maximum Hc is oriented along [100]opx and parallel to the minimum χ*, which reflects larger vortex nucleation fields when the applied field direction approaches the short axis. The maximum Hc is therefore orthogonal to the maximum Mrs/Ms, controlled by axis-aligned metastable single-domain states at zero field. The results emphasize that the nature of anisotropy in natural magnetite does not just depend on the particle orientations, but on the presence of different stable and metastable domain states, and the mechanism of magnetic switching between them. Magnetic modelling of natural magnetic particles is therefore a vital method to extract and process anisotropic hysteresis parameters directly from the primary remanence carriers.more » « less
