GRAITEC Group News
Construction of the building of the Jérôme Seydoux-Pathé Foundation: OMNIS Bâtiment comments...
Since 2006, OMNIS Bâtiment specializes in complex structures calculations or structures located in seismic areas. Its team of engineers, dedicated to design offices and construction companies, carry out building in-depth studies for conception and especially in the construction phase. Formerly equipped with Effel software, OMNIS Bâtiment now works with Advance Design, the worthy successor to Effel in terms of finite elements calculation.
Alexandre Jennan, Associate Engineer, discusses this major project.
What is the context of this project?
This building is intended to become the new headquarters of the Jérôme Seydoux-Pathé Foundation, located in the 13th district of Paris at 73, avenue des Gobelins. The opening has been planned for the second half of 2013. Designed by Renzo Piano, the building will maintain and highlight the facade sculpted by Rodin which will stand in front of a 5-story shell in the middle of a garden covered with a glass roof.
The building with an area of 2 200 m2 will include the offices of the Foundation, the archives, a documentation and research center, a DVD library, a space dedicated to exhibitions and a projection room.
The various stakeholders for studies of this project are:
- The architecture firm Renzo Piano Building Workshop (RPBW).
- VP&Green as project manager and engineering design.
- RBS design office for the design of the steel framework and the coordination of construction drawings.
- AD Structures office for all reinforced concrete design except those for the shell.
- Paul Sabia, a freelance draughtsman, who has produced the reinforcement drawings in 3D for the shell.
- OMNIS that have made the overall modeling and all calculations of reinforced concrete shell.
Can you give us the technical features?
The key feature of the project is its facade "3D concrete shell" with pronounced curves and shapes suggesting a connection to each neighboring building. The shell does not reach the ground level and is generally hanged by the intermediate levels. As far as we know, this entirely suspended reinforced concrete shell is the second one built in France after the one built for the Archipel Theater, also designed by our teams.
In this unique project, the shell with a thickness of 22 cm has a variable mechanical behavior. On a given stage, interior steel floors are in some places suspended by the shell periphery, which then behaves like a 3D curved deep beam... Up to reach areas where floor's cantilevers become stiff enough to support the shell. Then, according to the story we see the distribution vary between supporting areas and planned areas.
Overall, there are a hundred connection points between the shell and the steel floor beams, about a point every 5 m2 shell. They are the nodes of a kind of macro-mesh shell, which efforts allow to specify the overall behavior of the shell under various stresses, gravity and thermal strains. Some points have a positive reaction, others a negative and it seems to be highly variable. For the durability of the work, its good constructability (and demolition) we have made an iterative sort through all possible points of contact by setting a goal that everyone keeps a constant and consistent structural behavior to its environment: no significant variation between a constant point and the neighboring ones (limitation of internal stresses), constant sign of the reaction at a point whatever changes overloads interior floors (archives).
What was your approach to this project and what benefits do you get from the use of Advance Design?
There were three major steps in our approach as described below:
1/ Modeling 3D shell:
Initially, it was a "Rhino" project that was exported in DXF 3D and imported in Advance Design. Over 10,000 stitches (facets) were created with the DXF file as a medium in order to more closely model the complex geometry of the shell.
For time reasons, three engineers worked simultaneously on three different Advance Design models for modeling the shell. Using the functionality of import / export of structures in libraries, we were able to consolidate the work of these three engineers to get the entire 3D model of the shell. Basements and interior structures were then imported from an Arche Ossature model.
The flexibility and power of modeling, the ability to consolidate parts of the works have made possible this collaborative work.
2/ The requirements of the numerous intermediate stages
It was necessary to take into account the different intermediate stages corresponding to the steps of construction with temporary shoring and off-shoring per level. The constructive aim was to achieve a gradual deformation, first to control our final 3D geometry whose tolerance execution level was very fine, then to overcome sudden changes in structural behavior which can be detrimental to the building. The calculation goal was then to better follow the distribution of forces in the different stages, and the adequacy of reinforcement.
This represents a total of ten different models that interact with each other. So we got loads of some models to inject them in others, according to the stages of construction.
"For this, we have used and abused a particularly interesting feature of Advance Design which allows to import actions via a text file. This feature was very helpful and a significant time savings. "
We have also used the loads import function in a text file in order to integrate, into the Advance Design model, the complex load distribution of the canopy.
When setting the different models we have also used a lot the ability to update graphical views operating results. Here again, we gained a lot of time by automating repetitive tasks as the definition of operating scenarios is done only once and then automatically updated.
3/ Justification and cracking controlling in accordance with EC2
Another point that was important on this project focused on the control of cracking. We had on this project a maximum crack opening set at Wmax = 0.3mm. The control of cracking was performed by direct calculation of the crack opening and not by the use of more pessimistic tabulated values and resulting in a greater amount of reinforcement.
The accurate and efficient implementation of Eurocodes in Advance Design was very helpful on this point:
- We have justified the criteria for cracks opening directly exploiting Advance Design features:
- First calculation of theoretical reinforcement values to determine the level of cracking.
- Second calculation by imposing a real reinforcement in the planar elements (with reinforcement mini # HA10 e = 15cm per side) => we could use graphical regions giving values Wk (crack opening) and thus identify areas requiring additional reinforcement to meet the goal of cracking control (with consideration of creep).
- New calculations by imposing greater local real reinforcement in areas of high cracking level. Advance Design has therefore automatically recalculated the new corresponding cracks openings and we were able to justify the entire structure (with the actual reinforcement) in a few iterations.
Have you faced any difficulties on this project?
We have been able to run this complex project thanks to the flexibility of Advance Design and to the accurate Eurocodes implementation, especially the EC2 for the control of the cracking.
Generally, we consider the positive aspects by using Advance Design for this project.
Nevertheless, if we have to point out some points that could be improved in the software, we could mention:
- Improved function to hinge the surface elements on their contours.
- A better management of large files even if we went further this problem by splitting the project into several parts files sizes.
- The ability to directly impose a user creep coefficient on complex projects like this, even if the creep coefficient is calculated automatically based on user settings.
- The ability to store somewhere in the user text file the description of the context, assumptions and input characteristics of that file.
- The addition of a kind of assistance for the comparison of results from many consecutive calculations!