A structured approach to forensic structural investigations of concrete damages

Abstract

In the Netherlands, there are generally two categories of structural forensic investigations. The first category are extensive investigations with the goal to learn from failure by investigating all aspects: technical, human and organisational. The second category are investigations that focus mainly on technical causes and the investigation goal is often determining the failure cause including designing repair measures. For this category, little information is available on used investigation processes.

In literature, several investigation methodologies have been described. However, they often describe only a very general process or on the other hand very specific material related laboratory tests. Descriptions and recommended techniques on how to perform the proposed steps of an investigation are scarce. To a client ordering a damage investigation, the most important parts of the investigation are the conclusion and the recommendations. The client expects that the conclusion is correct. Therefore, it is important to ensure that the used investigation process results in a reliable (consistently good quality and able to be trusted) outcome. Basic knowledge of bias may help to indicate possible threats to the reliability of investigations. Techniques used to increase the validity and reliability of case study research can assist in developing strategies to achieve reliable forensic structural investigations. Therefore, the question of this research is: what is a reliable methodology to perform investigations into the technical cause of damages to concrete structures?

The research question has been answered by studying available literature on structural forensic investigation techniques, interviewing investigators, researching investigation techniques used for aerospace accident investigations and fire cause determination, studying human error theory and case study research and concrete damage mechanisms.

Based on the resulting information from studying all these disciplines the conclusion can be drawn that using a structured investigation process is essential. Therefore, this information has been used to develop an investigation methodology called the ‘investigation process model’. The model recommends the following phases: orientation, data collection, hypothesis generation, hypothesis analysis, conclusion, reporting and follow-up. These phases are split into individual steps that guide the investigator through the process. Each step contains suggestions for specific techniques varying from taking meaningful photographs on location to using a standard layout for reporting in order to execute the step properly. The investigator can select the most appropriate techniques for the project. Special attention has been paid to provide techniques to generate and analyse hypotheses in a reliable way. An example is the tool ‘Concrete Damage Handbook’ to assist the investigator in linking visual damages to possible causes. As a common thread through the model measures have been provided to limit the negative influence of bias (based on the Delft approach).

Validation of the model with 1) a test case based on a real damage file, 2) comparison with real investigation reports and 3) scientific papers proved that the model relates to the daily working practice and is usable, practical and functional within the scope of the defined criteria.