Non-destructive Estimation of Concrete Compressive Strength: Databases
收藏NIAID Data Ecosystem2026-05-02 收录
下载链接:
https://zenodo.org/record/14921018
下载链接
链接失效反馈官方服务:
资源简介:
Repository Description
This repository presents an extensive collection of published studies correlating non-destructive test (NDT) results with concrete compressive strength. The included NDT methods are:
Ultrasonic Pulse Velocity (UPV)
Rebound Hammer (RH)
SonReb (combined UPV and RH)
The dataset comprises 12,246 test results from 77 published studies, distributed as follows:
UPV: 4,846 tests from 66 studies
RH: 7,400 tests from 59 studies
SonReb: 2,735 tests from 41 studies
In each database, the final response variable is the concrete compressive strength, normalized to a 150×300 mm reference cylinder.
Database Structure
Each database contains approximately equal proportions of numerical and categorical variables:
UPV: 20 input variables
RH: 18 input variables
SonReb: 25 input variables
These variables cover material age, location, composition, NDT parameters, and testing procedures. Of the total dataset, 1,087 UPV tests, 2,052 RH tests, and 473 SonReb tests were conducted on in-situ structures, while the remainder were performed on laboratory-cast specimens. The datasets span material ages from the 1930s to freshly cast specimens, excluding tests on concrete less than one day old. Numerical data from visualized results in publications were extracted using an open-source computer vision-assisted software for high precision.
Project and Documentation
These databases were compiled as part of the EU-funded ReCreate project. The ReCreate project researches the process of reusing precast concrete elements through four real-life deconstruction and reuse pilots in Europe. The project aims to assess and improve the technical feasibility of a circular life cycle for structural precast elements across the entire value chain from disassembly, quality control, design for reuse and reconstruction.
A manuscript applying this dataset to empirical and machine learning models is under review and will be linked here upon publication.
For detailed descriptions of nomenclature, abbreviations, and assumptions, refer to the Database Guide PDF included in this repository.
创建时间:
2025-04-07



