Project SE-Lab
Structural Engineering Lab: Cloud/Grid-Based Virtual Laboratory for Non-Linear Probabilistic Structural Analysis
Structural Engineering Lab: Cloud/Grid-Based Virtual Laboratory for Non-Linear Probabilistic Structural Analysis
The Elbebridge Wittenberge bridge was chosen as the main use case and validation procedure in SE-LAB project.
The bridge is pilot project for a special composite roadway-slab, which is not covered by German codes. Because of that, large scale construction type tests had to be performed for the special admittance procedure. Three test structures had been built and tested under service and ultimate load conditions. The results of these real tests will be used to validate the results computed by the platform.
The Elbebrücke Wittenberge bridge will also be used for other aspects of the validation procedure considering the entire bridge.
The complexity of this project requires all main topics of the virtual laboratory, such as non-linear, probabilistic computations, that exceed the power of standard up-to date PCs and require grid/cloud computing.
The Bridge over the Elbe River is located in north-eastern Germany crossing the border of the states of Brandenburg and Saxony-Anhalt close to the city of Wittenberge. It will be part of the autobahn BAB A14 connecting the cities of Magdeburg and Schwerin. The bridge consists of a 696m long prestressed concrete approach bridge and a 412m long main bridge. The total length between the abutments is 1110.5m.
The main bridge is a 3-span continuous beam with spans of 126m - 160m - 126m. The cross-section is a 3-cell box girder with cantilever wings and a total width between the guardrails of 30.40m. The upper chord of the cross-section is wave-like curved and crates a varying height from 9.60m at the piers and 5.00m in midspan.
With a slenderness of 32, the superstructure requires a light but fatigue resistant roadway slab. These requirements are fulfilled by an “ortho-composite-slab” which consists of a reduced orthotropic steel deck and a 15cm thin in-situ concrete layer, connected by shear studs.
Roadway slab test in Labor
Key Validation Parameters
The exchange of BIM data via IFC from CAD to the ATENA software has to be checked. Furthermore BIM/IFC has to be extended with extra information for nonlinear and probabilistic analysis. The implementation of these additional parameters is a very important task that will have to be validated deeply.
The IFC-model of the entire bridge has a very high level of detail and is a challenge for validating the SE-Lab import/export interface. The program should allow the end user to choose the degree of detail how the bridge is modeled and determine macro sections of the structure.
For the efficient use of grid/cloud components, variations of the model used as the baseline have to be generated by the virtual laboratory. Although the immense increase of computational power compared to standard up-to-date PCs, the creation of model variations must not exceed the essentially needed amount, to reduce transferring and filtering of results.
Sending jobs to grid nodes and retrieve results at predefined points of interest is the other major task of the grid/cloud service.
The structural analysis solver will be validated by using the test data described in chapter 2.3. The focus will be on the nonlinear effects that have become apparent during the tests, as previously described.
Probabilistic approaches will also be validated by using the test data. As an example on Fig. 1 is seen that the green and red line are very close together, while the black line is further away. The expected distribution of these curves is much more consistent, than shown on the graph. Probabilistic considerations will show the exact origin of this deviation.
The front end of the virtual laboratory will be a web-based interface, where architects and structural engineers access the services of SE-Lab. It will have a user management layer, where users register, log in or get other information about the usage of the lab. In the next level BIM data will be uploaded, visualized and filters may be set. For offline pre- and post-processing input data and results have to be exchanged with the platform. Other layers for collaboration and communication between all actors and layers for graphical output of results e.g. as charts have to be created.