under pressure

INHALT In der diesjährigen Seminarwoche der Professur für CAAD wird eine selbstentwickelte materialeffiziente Metallkonstruktion einem Belastungstest in der HIF-Halle ausgesetzt. Vier Gruppen werden unter jeweils einer „Unterkonstruktion“ von 80x180x80 cm Grösse entwerfen und produzieren. Vom Entwurf, über Finite-Elemente-Simulationen bis hin zur computergesteuerten Produktion werden moderne Methoden des computergestützten Entwerfens und Produzierens praktisch angewendet.
Die erste Phase der Seminarwoche bildet eine Auseinandersetzung mit der Technologie der Blechbearbeitung über Entwurf hin zur Blechkonstruktion. Es werden aktuelle Produktionsmöglichkeiten präsentiert und von den Teilnehmerinnen in die Projekte integriert. Es entsteht eine digitale „Produktionskette Blech“.
Die zweite Phase widmet sich der Umsetzung und Realisation des Konzeptes. Im Einsatz kommen computergesteuerte Werkzeuge, welche state-of-the-art der CNC-Metallverarbeitung sind (Lasern und Stanzen, pneumatisches Biegen, Hochfrequenz- und Punktschweissen).
In der dritten Phase des Kurses finden Belastungsproben der Konstruktionen statt. Die Simulationen, die mit einem CAD-Programm, das Finitenelementen rechnet, erzielt werden, werden mit realen pneumatischen Belastungsproben verglichen.
Die Konstruktion, die das beste Verhältnis zwischen Gewicht und Tragfähigkeit erzielt hat, hat gewonnen.

Professur für CAAD Ludger Hovestadt: Prof. Dr.Ludger Hovestadt, David Sekanina, Oskar Zieta
Institut für Baustatik und Konstruktion, Prof. Dr. Mario Fontana

Under Pressure: Digital production and optimization of sheet metal structures

Abstract:

The advancements of parametric 3D modelers and finite element analysis applications, paired with the progress in CNC manufacturing, empower today’s architect building unique, complex and structurally sane sheet metal structures in a short time. To prove this point a one week workshop was held where the students were asked to build a sheet metal structure as stable and lightweight as possible, in the given dimension of 800*1800*800mm, using only 1mm thick steel. By using a parametric 3D modeler with integrated sheet metal functions, the structure could be modeled in its final bent and assembled state. Each component could be flattened directly in the software checking for overlaps. Using a top-down construction methodology, where new parts depend on dimensions and constraints of existing parts, the whole assembly updated automatically, when the dimension of one part were changed, allowing to change the statics of the whole assembly with a few clicks. With the finished assembly, the optimization cycle would start using finite element analysis (FEA) software. The FEA software used for the first tests lacked the nonlinear buckling analysis functions, essential for this task. We had to switch to a more powerful FEA application. The results of these virtual stress tests led to changes to the original structure, trying to distribute the stresses evenly. Now an optimization cycle would start where newly changed structure would be tested again until the buckling stresses were minimized. All parts were modeled as sheet metal parts in 3D. This enables the user to flatten the parts virtually and using the flat contours directly to create the 2D data that was sent to the CNC laser cutter. The precision of today’s CNC laser cutters and bending machines leads to highly precise bent sheet metal parts, simplifying and speeding up the assembly considerably. Very tight tolerances can be maintained, reducing structural instabilities due to imprecision. Various assembly methods were used in our five structures that were built. Pop rivets, spot welds, bolts and linear welds. Finally the constructions were mounted on a stress-strain bench. As the hydraulic cylinder could only pull, not push, the structures were mounted high up on columns with their ‘feet’ loosely attached to them and then pulled down from the center. Gradually the pulling force was elevated and the resulting displacements were measured automatically, resulting in a flow of data which then could be translated into a graph plotting the stress-displacement curve.

To our and the professor’s astonishment the structures were far more stable than we thought. Most of us initially thought that the structure would support a load of roughly 1300 to 1500 Newton. But even the weakest structure didn’t fail before applying a load of 6000 Newton. The strongest structure even supported a load of 37 kN.

Conclusion:

While in other industries sheet metal constructions are widely used, architects still fear of using them as a supportive structure. With this workshop we tried to show that within digital production chain a complex and statically safe structure can be built out of thin sheet metal parts in a short time.

Film: