------------------------------------------------------------------
Name: Mechanical Properties of Normal Concrete
------------------------------------------------------------------
Data type: Real and multivariable data
------------------------------------------------------------------
Associated tasks: Regression
------------------------------------------------------------------
Authors:

Farzad Rezazadeh (1), Axel Dürrbaum (1), Amin Abrishambaf (2), Gregor Zimmermann (3), and Andreas Kroll (1)

(1) University of Kassel, Faculty of Mechanical Engineering, Department of Measurement and Control
(2) QuantumFusion GmbH, Kassel, Germany
(3) MAITERIA UG, Kassel, Germany
------------------------------------------------------------------
Abstract:

Normal concrete is the most widely used form of concrete, and its mechanical properties can vary due to variations in raw-material quality, dosage errors, and changes in material storage, mixing, and curing conditions, even when a fixed reference mix design is used. 
This variability constitutes a reproducibility challenge for concrete production under fixed formulations.
This dataset examines the effects of variations in raw-material condition and process parameters on the mechanical properties of normal concrete produced from a base mix design targeting a 65 MPa compressive strength.
The dataset comprises 32 systematically designed experiments. 
Compressive strength was measured at 1 day (24 hours), 7 days, and 28 days after mixing. Where available, the reported values represent the average of three specimens per experiment.
In addition, five fresh-state properties were measured immediately after each mixing process (temperature, electrical conductivity sensor reading, slump-flow, V-funnel flow time, and air content).
All experiments were conducted in the laboratory of G.tecz Engineering GmbH under controlled conditions using the same mixer, mixing tools, and personnel.
The dataset provides high-dimensional experimental data with a limited number of observations and is suitable for developing and evaluating regression models in sparse scenarios.
------------------------------------------------------------------
Data characteristics:

All experiments were performed with the same mixing equipment under controlled environmental conditions for both material storage and production, thus minimizing seasonal or external variations. 
The mixer chamber was maintained at approximately 20 °C, and this temperature was verified before each experiment to ensure consistent mixing conditions. 
The experiments were conducted following a structured experimental plan (Taguchi orthogonal array L32) and the resulting data have not been scaled or normalized (raw form). 
Missing values may appear as “NM” and/or empty cells. Missing values occur not only in fresh-state properties but also in some compressive strength outputs. 
In particular, “NM” marks entries where a fresh concrete property could not be measured due to insufficient workability. 

Data summary:

•	Number of experiments (observations): 32
•	Number of inputs: 16 (13 quantitative input variables, 3 qualitative input variables)
•	Number of fresh concrete properties: 5 (all quantitative fresh-state measurements)
•	Number of outputs (mechanical properties): 3 (Compressive strengths at 1 day, 7 days, and 28 days after mixing)
------------------------------------------------------------------
Variable information:

•	Component 1: Experiment number
•       Component 2: Cement reactivity [Qualitative: 1 or 2] ––> 1: Lower reactivity cement. 2: Higher reactivity cement.
•	Component 3: Graphite in kg/(15 L)
•	Component 4: Ingredient moisture in kg/(15 L)
•	Component 5: Ingredient temperature in °C
•	Component 6: Sand (0–4 mm) in kg/(15 L)
•	Component 7: Fine aggregate (2–8 mm) in kg/(15 L)
•	Component 8: Coarse aggregate (8–16 mm) in kg/(15 L)
•	Component 9: Filler I in kg/(15 L)
•	Component 10: Filler II in kg/(15 L)
•	Component 11: Superplasticizer in kg/(15 L)
•	Component 12: Mixer speed in rad/s
•	Component 13: Mixing duration in s
•	Component 14: Curing temperature day 1 in °C
•	Component 15: Curing class day 1 [Qualitative: 1 or 2] ––> 1: Storing the specimens in a humidity-controlled cabinet at 90% relative humidity. 2: Covering the specimens with plastic film.
•	Component 16: Curing temperature days 2–28 in °C
•	Component 17: Curing class days 2–28 [Qualitative: 1 or 2] ––> 1: Maintaining the specimens in plastic film (sealed curing). 2: Submerging the specimens in water (water curing).
•	Component 18: Temperature (fresh concrete) in °C
•	Component 19: Electrical conductivity sensor signal (fresh concrete), ADC counts ––> Raw ADC output code from the fresh-concrete conductivity measurement circuit (Vref = 5 V); values can be converted to the sensor output voltage using (X × 5/2048) and should be interpreted as an uncalibrated proxy for electrical conductivity rather than an absolute conductivity value.
•	Component 20: Air content (fresh concrete) in %
•	Component 21: Slump flow (fresh concrete - t=120) in mm ––> Slump-flow diameter measured at 120 s after lifting the cone.
•	Component 22: V-funnel runtime (fresh concrete) in s
•	Component 23: Compressive strength day 1 in MPa
•	Component 24: Compressive strength day 7 in MPa
•	Component 25: Compressive strength day 28 in MPa
------------------------------------------------------------------
NOTE: Reuse of this dataset is unrestricted with retention of the original copyright notice.
------------------------------------------------------------------