Large testing facilities are used in combination with advanced simulation tools for the assessment of structures subjected to a variety of loadings. In earthquake engineering this is the case when testing full scale structures or structural elements. The numerical simulations are used for verification of the tests, or as integral parts of the tests themselves, as e.g. in the sub- structuring method.
Experimental testing plays an important role in the development and evolvement of design rules and the verification of new products and methods in the construction sector. The experimental campaigns often very costly and involve a large number of participants. They yield input to advanced computer simulations run simultaneously with or after the experimental campaigns, for the verification of the tests themselves.
As the laboratories are expensive to build and costly to run it would be an advantage to increase the number of researchers and projects who can benefit from the individual experiments. One way to achieve this is to make the various test campaigns more visible and accessible for the research society in order to benefit form other ongoing research, to share test facility resources and to collaborate more interactively in ongoing research projects.
In order to facilitate collaboration with external partners all over the World, a system for Distributed Laboratory (DL) is being installed at ELSA. The system allows external colleagues to follow the experiments via the Internet, and offers the possibility to display measured signal values in real time, selected among a list, and to view video of the experiment online. Thus, external collaborators can contribute to the decision making and conduction of the experiment in real time.
The main objectives of the Distributed:
- To make the tests more visible in that more people should be allowed to follow the tests. This requires a telepresence infrastructure with the capacity to deliver enough information without disturbing the running of the test.
- To make the tests more interactive, though the control of the test remains in last instance the responsibility of the laboratory running it. Basic interactivity is provided when a remote user is able to specify the next maximum seismic acceleration in a series of test of growing intensity. Larger interactivity is required when performing a distributed Pseudo Dynamic test with sub-structuring.
- To improve the documentation of the tests. Usually, only the group of people present during the test knows what really happened. Sometimes some minor problems which are not mentioned afterwards could have an incidence on the modelling. When the test goes wrong with respect to its objectives, it could still be useful for modelling.
- To make the tests more retrievable in the meaning that the tests that are archived should be as widely used as possible. They should be stored in a meaningful way, easily accessed and retrievable. This last point is crucial, and it is not expected to be solved at the level of a laboratory. What is needed is a distribution over a community.
All the points briefly outlined above serve to enlarge the participation to the test at the level of its preparation, running, analysis, exploitation and simulation. This is clearly a distributed effort and the result could be called a distributed laboratory.
The first step to distribute a test is to distribute the results. Common database, common data formats are currently discussed at a world level within an initiative of NEES, the US Network for Earthquake Engineering Simulation, which includes a hardwired network of 15 laboratories, distributed over the US country.
The second step is to ensure telepresence; that is, the possibility for a remote user to participate to a test. The user is no longer in the laboratory but is provided with a sufficient amount of key information in order to permit him to be an active player of the test. Telepresence allows for more end-users and increasing the visibility of the test. Telepresence is saving the time (and cost) of travelling, giving more flexibility in the test planning, limiting the stress and thus increasing the quality of the test.
The third step is to distribute the test over a network of geographically distant laboratory. The idea is to combine the specific capabilities of different laboratory to the realization of a joint test where the various facilities involved are working on-line and exchange information using internet. It is evident that pseudo-dynamic testing with sub-structuring is the first approach to investigate for distributing a test.