Start Date

2016 12:00 AM

Description

There are many commercial ultrasonic tomography devices available for use in nondestructive evaluation (NDE) of reinforced concrete structures. These devices operate by using an array of transducers to transmit and receive ultrasonic data. Time-staggered ultrasonic impulses are released at each transmitting transducer through a concrete medium, after which the shear wave reflections are recorded at all receiving transducers. This ultrasonic data representing the received impulse responses between all unique transducer pairings are collected by the devices and often stored in a proprietary format. Because of this, both academic and industrial researchers are prevented from using the collected data for NDE algorithm development and testing.

Thus, there is a motivation for a proper sharable format to exist such that all collected proprietary data sets for a specific concrete specimen can be converted, organized, and stored with relative metadata for end user NDE algorithmic image reconstruction and research. The conversion would allow a party with rights and knowledge of the proprietary format to repackage the ultrasonic data to a more standard data type, such as a numerical array of integers or floats, which is sharable with those without similar rights or access. It also follows that a proper framework must exist for effortless reading and writing of the new data format.

A concrete NDE framework that is independent of proprietary format limitations has been developed for the use of data conversion and storage from the proprietary formats, the reading and analysis of converted data, and the testing of current and new NDE algorithms. It is built upon the popular and open hierarchical file format HDF5, which was chosen due to its portability, extensibility, and the availability of high/low level access functions in many programming language libraries. The choice of HDF5 allows for concrete specimen-specific constants, variables, and descriptors to be stored as HDF5 attributes alongside ultrasonic tomographic data sets, with different attributes associating with different HDF5 branches at different depth levels.

The framework allows one to write multiple proprietary files from different data sets in a concrete specimen to the hierarchically organized HDF5 format with groupings intelligently made for the organization of ultrasonic data as well as all related attributes. Attribute data for a specific data set can be read to override default constants and parameters used in a particular NDE reconstruction algorithm, allowing for each reconstruction to be tailored to a particular ultrasonic data set and mitigating error introduced by hard coded generic values applied to all data sets. This supports the research of concrete specimen with a variety of different forms of degradation or other variable conditions.

The work is funded by the U.S. Department of Energy’s office of Nuclear Energy under the Light Water Reactor Sustainability (LWRS) program. The authors would like to acknowledge generous support of the U.S. Department of Energy.

Language

en

File Format

application/pdf

Share

COinS
 
Jan 1st, 12:00 AM

An HDF5 Based Framework for the Distribution and Analysis of Ultrasonic Concrete Data

There are many commercial ultrasonic tomography devices available for use in nondestructive evaluation (NDE) of reinforced concrete structures. These devices operate by using an array of transducers to transmit and receive ultrasonic data. Time-staggered ultrasonic impulses are released at each transmitting transducer through a concrete medium, after which the shear wave reflections are recorded at all receiving transducers. This ultrasonic data representing the received impulse responses between all unique transducer pairings are collected by the devices and often stored in a proprietary format. Because of this, both academic and industrial researchers are prevented from using the collected data for NDE algorithm development and testing.

Thus, there is a motivation for a proper sharable format to exist such that all collected proprietary data sets for a specific concrete specimen can be converted, organized, and stored with relative metadata for end user NDE algorithmic image reconstruction and research. The conversion would allow a party with rights and knowledge of the proprietary format to repackage the ultrasonic data to a more standard data type, such as a numerical array of integers or floats, which is sharable with those without similar rights or access. It also follows that a proper framework must exist for effortless reading and writing of the new data format.

A concrete NDE framework that is independent of proprietary format limitations has been developed for the use of data conversion and storage from the proprietary formats, the reading and analysis of converted data, and the testing of current and new NDE algorithms. It is built upon the popular and open hierarchical file format HDF5, which was chosen due to its portability, extensibility, and the availability of high/low level access functions in many programming language libraries. The choice of HDF5 allows for concrete specimen-specific constants, variables, and descriptors to be stored as HDF5 attributes alongside ultrasonic tomographic data sets, with different attributes associating with different HDF5 branches at different depth levels.

The framework allows one to write multiple proprietary files from different data sets in a concrete specimen to the hierarchically organized HDF5 format with groupings intelligently made for the organization of ultrasonic data as well as all related attributes. Attribute data for a specific data set can be read to override default constants and parameters used in a particular NDE reconstruction algorithm, allowing for each reconstruction to be tailored to a particular ultrasonic data set and mitigating error introduced by hard coded generic values applied to all data sets. This supports the research of concrete specimen with a variety of different forms of degradation or other variable conditions.

The work is funded by the U.S. Department of Energy’s office of Nuclear Energy under the Light Water Reactor Sustainability (LWRS) program. The authors would like to acknowledge generous support of the U.S. Department of Energy.