TO TOP

Quality assured flow production of lightweight UHSC rod elements using artificial neural networks

Applicants

Prof. Dr.-Ing. Ludger Lohaus Leibniz Universität Hannover, Fakultät für Bauingenieurwesen und Geodäsie, Institut für Baustoffe

Prof. Dr.-Ing. habil. Raimund Rolfes Leibniz Universität Hannover, Fakultät für Bauingenieurwesen und Geodäsie, Institut für Statik und Dynamik

Scientific staff

Dipl.-Ing. Jan Markowski Leibniz Universität Hannover, Fakultät für Bauingenieurwesen und Geodäsie, Institut für Baustoffe

Nikolai Penner, M. Sc. Leibniz Universität Hannover, Fakultät für Bauingenieurwesen und Geodäsie, Institut für Statik und Dynamik

Jicheng Yuan, M. Sc. Leibniz Universität Hannover, Fakultät für Bauingenieurwesen und Geodäsie, Institut für Statik und Dynamik

Project description

Compared to automated production of other industrial sectors, concrete construction is still significantly characterised by craft activities. The target properties of components will only be reached on site. They are therefore vulnerable to the specific environmental conditions, which can hardly be influenced. This situation leads to imprecision and uncertainty during construction works, which results in an uneconomic material usage and disturbances in the construction process. The results are long construction periods and waiting times.
With the concept “Individuality on the whole – similarity on the small” the DFG Priority Program shall investigate fundamentally new construction methods, which aim at a disruptive change in construction.
Together with the project partner Institute of Structural Analysis, the Institute of Building Materials Science is researching a novel manufacturing process for components made of ultra-high-strength concrete with a reinforcement of steel sheet and carbon fibres. An innovative extrusion process is used to produce rod-shaped components with a core of ultra-high strength concrete. They are reinforced by a combination of carbon fiber reinforced plastic and sheet steel. A sensor concept is being developed which is capable of monitoring the components "from the birth of the component". Various heterogeneous measurement data are used to control and monitor the extrusion process by means of an artificial neural network, so that a consistently high quality of the components can be guaranteed.

Poster on the project contents

Figure 1: Schematic representation of the component and the extrusion process
Figure 1: Schematic representation of the component and the extrusion process

Publications

[3] Markowski, J., Meyer, M., Haist, M., & Lohaus, L.
Duktilitätssteigernde Bewehrungssysteme für fließgefertigte Stabelemente aus UHFB.
Beton‐und Stahlbetonbau, 117(12), 998-1007.
https://doi.org/10.1002/best.202200062

[2] Tritschel, F. F.; Markowski, J.; Penner, N.; Rolfes, R.; Lohaus, L.; Haist, M.
KI-gestützte Qualitätssicherung für die Fließfertigung von UHFB-Stabelementen
Beton- und Stahlbetonbau 116, Sonderheft Schneller bauen S2, September 2021, S. 34–41.   
(https://doi.org/10.1002/best.202100052)

[1] Lohaus, L.; Rolfes, R.:
Qualitätsgesicherte Fließfertigung leichter UHFB-Stabelemente mittels Künstlicher Neuronaler Netze.
In: BetonWerk International Nr. 2, 2021, S. 18
Link zum Artikel

 

Supervised theses

2022

Forto, D.
Ultra-hochfester Extrusionsbeton – Schaltbare Eigenschaften
Bachelorarbeit, Leibniz Universität Hannover, Fakultät für Bauingenieurwesen und Geodäsie, Institut für Baustoffe, Betreuer: Markowski, J.

2020

Müller, M.
Strangpressen von Beton Hohlprofilen
Masterarbeit, Leibniz Universität Hannover, Fakultät für Bauingenieurwesen und Geodäsie, Institut für Baustoffe, Betreuer: Markowski, J.