To investigate the mechanisms involved during semi-brittle flow, we deformed Carrara marble over a confining pressure (Pc) range of 10-300 MPa and room temperature to ≈10% strain. We apply triaxial loading to intact Carrara marble and collect mechanical, ultrasound pulsing, and acoustic emission (AE) data during pressurization and deformation stages. The pulsing and AE waveforms are recorded using a pair of piezoelectric sensors. At lower Pc, microcracking is the dominant deformation mechanism, whereas at higher Pc, crystal-plastic mechanisms such as twinning and dislocation glide are favored. These changes in the activity of defect populations are manifested in changes in mechanical properties, velocity variations, and AE characteristics. Samples at higher Pc exhibit higher strength and require more work for fault-development. Transition from localized faulting to distributed barreling is observed between 50 and 100 MPa Pc. We track precise velocity variations from the pulsing waveforms using correlation-based methods. During the pressurization stage, the velocity increases logarithmically with Pc between 0-100 MPa, followed by a linear increase at higher pressures. During the deformation stage, the compressional wave velocity initially increases before the yield point due to closing of crevices, and then decreases exponentially after the yield point. The rate of this velocity decay is smaller as Pc increases, owing to reduced microcracking with very little change at Pc ≥ 200 MPa. AE data show that individual defect types emit characteristic patterns. Twinning produces repetitive patterns of low amplitude, short signals localized in frequency space whereas microcracks are more energetic, emit over a much broader frequency range, and show more variation in signal shape and duration. The AE spectra shift from ≈ 500 kHz to ≈15 MHz mean frequency as Pc increases, which is associated with increasing twinning activity. This acoustic data agree with microstructural observations of microcracks and crystal-plastic deformation in the samples. By joint-analyzing the stress-strain and velocity evolutions with AE observations, we obtain detailed changes in the micro-mechanisms accommodating strain in the Carrara marble and constrain the deformation modes as it goes through the brittle-plastic transition.

Laboratory data are essential for testing and refining of theories and models in Earth sciences. Recent developments in data mining techniques and machine learning have made it feasible to utilize and digest large amounts of information; yet such data must be initially prepared and structured in a meaningful way. Recognizing the potential and challenges of data access, multiple efforts are underway in the development of digital data repositories (Strabospot, Epos). Currently most information in experimental Geophysics is not accessible in digital, searchable form. Such information may include: equipment capabilities and configurations, original and edited experimental data, laboratory calibrations as well as information regarding testing protocols and procedures. The LAPS project aims to design and develop resources to facilitate data workflow and access. It specifically focuses on the needs of laboratory researchers, students and managers to prepare data for use in digital data repositories. As laboratories use a wide variety of hardware and software solutions to acquire and process data, we focus our efforts on the development of web based tools that do not require specific local infrastructure and software. One of the main objectives of LAPS is to establish a coherent and effective way to describe equipment and experiments across a variety of testing rigs and devices. To simplify the workflow we are proposing a combination of selecting pre-configured equipment and experimental profiles and manual data entry via web form. A completed test protocol containing all experimental metadata may then be saved locally (e.g., as JSON file) or (optionally) to a database. Such test protocols can be re-loaded and modified as needed. To complete the workflow of a successful test, a link to original or processed data files may be inserted. The data file layout can be defined in the experimental profile. Upload to a digital data repository is optional but by adhering to the proposed Strabo/Epos data model it will not require additional input. While the web based workflow will be accessible to all users, we also recognize the need to integrate the system into specific work flow solutions in rock deformation laboratories. We therefore provide the framework that simplifies local data management systems and analytical applications.