Data from uniaxial compressive testing of laboratory-made granular ice

Uniaxial compressive tests of laboratory-made granular fresh-water ice were conducted in a cold room in the ductile and brittle strain rate range at -10.83°C ±0.74°C. Ice specimens with a length to diameter ratio of 2.5 showing brittle behavior failed by axial splitting. With the Instron Labtronic 8800, the operator controlled the tests at a frequency of 4,000 Hz. The data acquisition rate was 25,000 Hz, and for faster experiments, 100,000 Hz. The operator controlled on a random basis the hydraulic cylinder by either the cylinder displacement or the specimen displacement. Increasing as well as constant and decreasing compression strength trends with increasing strain rates could be shown in the past. The data presented here show a lower compressive strength at strain rates higher than 4*10−3 s−1. The data consist of the time history of the specimen and cylinder displacement measurement (in mm), and the force measurement (in kN). The data is available as a separate .xlsx file for each test performed. In total, 123 tests were performed. If the test was performed with a 10 mm gap, the label ends with a ‘g’. The abbreviations are separated with an underscore. The data provided here can be used to validate ice-material models or for ice-testing databases for machine learning purposes.


a b s t r a c t
Uniaxial compressive tests of laboratory-made granular freshwater ice were conducted in a cold room in the ductile and brittle strain rate range at -10.83 °C ±0.74 °C. Ice specimens with a length to diameter ratio of 2.5 showing brittle behavior failed by axial splitting. With the Instron Labtronic 8800 , the operator controlled the tests at a frequency of 4,0 0 0 Hz. The data acquisition rate was 25,0 0 0 Hz, and for faster experiments, 10 0,0 0 0 Hz. The operator controlled on a random basis the hydraulic cylinder by either the cylinder displacement or the specimen displacement. Increasing as well as constant and decreasing compression strength trends with increasing strain rates could be shown in the past. The data presented here show a lower compressive strength at strain rates higher than 4 * 10 −3 s −1 . The data consist of the time history of the specimen and cylinder displacement measurement (in mm), and the force measurement (in kN). The data is available as a separate .xlsx file for each test performed. In total, 123 tests were performed. If the test was performed with a 10 mm gap, the label ends with a ' g' . The abbreviations are separated with an underscore. The data provided here can be used to validate ice-material models or for ice-testing databases for machine learning purposes.

Value of the Data
• The dataset is relevant as it is a complete presentation of uniaxial compression tests of granular ice at a high degree of detail. • Experimenters can benefit from the data, as it shows influences of the testing properties on the uniaxial compression tests of ice over a wide range of testing velocities/strain rates, which was not provided before at this degree of detail. • Data scientists with a material modeling background can benefit from the detailed data, as it shows influences of the strain rate on the uniaxial compressive strength of ice. • The presented data can be used to validate numerical ice material models or for data-driven analyses and modeling of uniaxial compression tests of ice.

Data Description
The data being shared describe the uniaxial compression properties of laboratory-made granular ice. Tables 1-4 show the testing settings: Set velocity for the compression test, if the operator controlled the test with the cylinder displacement (CDC) or with the specimen displacement (SDC), the PID controller settings, the data acquisition rate and the repository file name where the time history of the force and displacement measurement is stored. In the post-process of the data acquisition, the data was cut, and for tests with set velocities between 0.01 mm/s to 0.1 mm/s the sample rate was decreased by a factor of 100 to decrease the file size. The start point of the cutting process was a measured force of 1 kN, and the endpoint was either the maximum measured force for tests showing brittle behavior or a measured displacement around 10 mm for tests showing ductile behavior. The zero point of the time and displacement measurement was shifted to the start point, also defined as the point of contact.     Ice specimens with a length to diameter ratio of 2.5 showing brittle behavior failed by axial splitting. The data presented here show a lower compressive strength from strain rates higher than 4 * 10 −3 s −1 , compared to the other studies. The strength values in the brittle range are far below the strength values of Shazly et al. [1] and Jones [2] . The difference to Jones may be caused due to the different L/D ratio (2.08) and the columnar grain type, whereby Shazly has investigated single-and polycrystalline ice at much higher strain rates. Lian et al. [3] tested cubic, columnar ice specimens with a L/D ratio of 1 and presented the results over the apparent strain rate. Comparing different experimental ice compression studies is challenging as testing properties, ice specimen geometries, and grain types differ, and the indication of the fracture mode is sometimes lacking.

Experimental Design, Materials and Methods
The experimental design included the PID controller settings and the displacement control type and consisted of determining the compressive strength at high rates. The rate limitation was the Schenck Hydropuls-longitudinal PL 160 cylinder with its force capability of nominal 160 kN and a maximal velocity of 1 m/s. The hydraulic cylinder can be controlled either by the piston displacement (build-in inductive displacement sensor, CDC) or by the specimen displacement (SDC). The specimen displacement controlled tests neglect the effect of the test rig stiffness. All tests are performed at TUHH in the mechanical laboratory of the Institute for Ship Structural Design and Analysis.

Specimen production
Cylindrical ice specimens with a diameter of 99.4 mm and a total length of 250 mm were produced in an approximately -10 °C cold refrigerated container.
PVC-U tubes, commercially available crushed ice, distilled water, steel plates, and insulating material were needed to make granular ice specimens. The distilled water had a mean acidity of 8.6 pH ±1 pH and a mean electrical conductivity of 7.5 μS/cm ± 5.8 μS/cm measured at a random basis at temperatures of 18.8 °C ± 3.6 °C.
The tubes were cut into the desired length, including a minimum addition of 30 mm regarding the cutting process after the specimen removal, and glued to the steel plate on one side. The steel plate formed the base and ensured good heat conduction. In the refrigerated container, two-thirds of crushed ice and one-third of distilled water was added to the tube, and the top of the tube was covered with insulation material. After freezing for at least two days, the ice specimen was removed from the tube by heating it with a hot air gun up (about 2,0 0 0 W) for a short time. After removal, the ice specimens were stored in the refrigerated container and were cut at both ends into the desired length with a band saw located in the refrigerated container. This process ensured consistency in the grain structure and shape of the specimens. Fig. 2 shows a thin section of an exemplary specimen (grain size approximately 1-10 mm determined by manual counting).
Herrnring et al. [4 , 5] already used the ice production process successfully, based on Gudimetla [6] , but in contrast, commercial crushed ice instead of self-made crushed ice was used.

Setup and methods
The Schenk Hydropuls-Longitudinal PL 160 cylinder , the position of the load cell, and the position of the potentiometric displacement sensor burster 8718-500 for the specimen displacement measurement are shown in Fig. 3 .  With the Instron Labtronic 8800 , the operator controlled the tests at a frequency of 4,0 0 0 Hz. The measurements are carried out with the Gantner Q.raxx-station 101 T measuring amplifier. The displacement, the force and the temperature are measured with the Q.raxx A101 module. The data acquisition rate was 25,0 0 0 Hz for test series 1 -3 and 10 0,0 0 0 Hz for tests series 4, see also Table 1 to Table 4 . Two operators were needed to perform the tests. One operator controlled the hydraulic cylinder, and the other operator started the measurement and prepared the measurement environment. The measurement environment had to be cleaned before each test, and the operator had to place the specimen centrally.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data Availability
Data from uniaxial compressive testing of laboratory-made granular ice (Original data) (Mendeley Data).