University of Rijeka, Croatia

Topics of structural dynamics and earthquakeengineering are covered in most, if not all civil engineering programs. However, not all universities are equipped to perform extensive experiments and testing in this area, leaving the teaching and research at a theoretical and simulation level.

The Faculty of Civil Engineering at the University of Rijeka is fortunate in that sense. Their lab can boast about not one, but two bi-directional shake table platforms. With this unique setup, the research team of Professors Nenad Bicanic and Gordan Jelenic was able to take their research to the next level.

Challenge

More Insights into Dynamic Response of Discontinuous Systems

Many structures are built using multiple blocks of material, with gaps and clearances of simple or rather complex shapes in between. The spaces between the blocks allow for limited sliding and rocking during the dynamic response to excitation, for example during an earthquake. Civil engineering classifies these structures as discontinuous. Stone or brick walls are typical examples of such structures, but so are graphite cores used in nuclear power plants, where gaps help avoid extensive thermal stresses.

Often, they represent a vital safety-critical component of the entire structural system. Being able to predict the behavior of discontinuous systems under both static and dynamic conditions more accurately would help design safer structures and develop safety assessment procedures.

But to assess or characterize a potential sensitivity of discontinuous systems to dynamic excitation is extremely difficult. Computational simulations typically rely on methods developed for continuous systems and extending them to multi-block assemblies, which is not adequate. The scalability of experiments with discontinuous structures presents another challenge.

With that in mind, and with the unique Shake Table III setup in the lab, the team at the University of Rijeka set off to get more insights into thedynamic response of the discontinuous systems.

Solution

What Makes the Blocks Collapse

The team studied the conditions for over-turning of a single block column and multi-block assembly excited by a ground motion, and the effect of the surface stiffness on the overturning conditions.

To do that, they placed the block column and multi-block structures on 3D-printed baseplates emulating softer or harder surfaces and excited them by running sine waves of different frequencies on the Shake Table III. Through a set of experiments, the team obtained the block displacement time histories and identified the ground motion function resulting in the complete or partial overturning of the structures.

The initial results suggested a broad agreement between experiments and simulations of the block structures’ qualitative behavior. Further research would be required to validate simulations of the dissipation mechanisms on the inter-block, as well as block-base interfaces, and to investigate suitable scaling parameters for practical assessment of real-life structures.

How Blocks Self-organize
In the second project, the researchers studied how blocks placed in a constrained space, and exposed to horizontal harmonic excitation, form patterns. Their goal was to find out if, and under what conditions, the blocks would converge to a repeatable and predictable motion.

The team performed a series of experiments with four or eight blocks confined in a basin, capturing the displacement of the blocks with an optical measuring system. To ensure minimal friction between the blocks and the basin floor, the blocks were placed on top of an aluminum beads layer.

In these experiments, the team employed two Shake Tables III. Running them synchronously allowed the researchers to examine the effect of the same excitation on different structures, or on the same structures, but with different damping measures.

The results suggested that for specific conditions, the assembly response patterns of the blocks were repeatable, although minor changes in the excitation frequency and amplitude led to different response patterns, where the multiblock assembly responded as a single block, or the blocks separated and came together in repeatable fashion.

Behaviour of Structures under Multi-support Excitation

Multi-block structures respond differently to ground excitation. Image courtesy of the University of Rijeka

Large structures, such as long bridges, present another challenge in earthquake engineering. Because of the greater distance between the supports, the excitation each support is subjected to may be quite different. This may result in structural failure due to the excessive relative displacement between the supports.

Current literature offers theoretically-based procedures for calculating dynamic response of structures subjected to multi-support excitation, but the experimental validation is not well documented. Such experiments would require a setup capable of giving different base displacement functions to each support of the structure at the same time.

The lab at the University of Rijeka, with two bi-directional Shake Tables, was well-positioned to take on this research challenge. The research team measured the horizontal deflection of a freely supported beam and a cantilever beam with additional discrete masses, subjected to synchronous and asynchronous, as well as uniform and non-uniform harmonic excitations. They compared the results with the analytical and numerical approaches to confirm the theoretical predictions. To measure the dynamic response of a long structure to excitation, the research team at the University of Rijeka used a 2 meters long wooden beam. Each of the beam supports was fixed to a separate Shake Table. The dual Shake Table setup allowed to apply excitations with the synchronous signal, or a signal with some delay. The team used the 3D non-contact optical displacement and deformation measurement system to study the horizontal deflection of the beam.

Result

Transforming the Students’ Experience

Using the AELabs in the course for two semesters did not allow Dr. Trescases to quantify their impact on the learning outcomes, but he considers the students’ experience in the lab as truly transformative. The change of attitude was almost immediate, with student engagement positively influenced by hands-on interactions with real-world circuit applications. For example, students were encouraged to bring their smartphones to the lab and play their favorite music through the amplifier circuit on the BJT experiment board. By modifying the configuration of the amplifier and they could hear the resulting changes in the audio output through their headphones.

Without AELabs, the chance students would graduate from the Electrical Engineering program without ever having seen or interacted with a surface-mount component was real. For Dr. Trescases that would mean he failed the students as an instructor. AElabs taught students how PCBs work, how devices are connected, and how the components are mounted, all of which is extremely useful.

The AELabs design earned the student team the University of Toronto’s 2013 Gordon R. Slemon Design Award, honoring the most imaginative design project, product design, and execution. After graduating from the Electrical Engineering program, the students from the AELabs design team founded the startup Illuster Technologies, with the aim to commercialize the product. The feature article on the AELabs, published in the ECE’s magazine Annum 2014, caught an eye of Dr. Jacob Apkarian, renowned modern control, and mechatronics pioneer, Quanser’s founder and one of Dr. Sedra’s students in 1970s. Dr. Apkarian recognized that AELabs was consistent with the core philosophy of all Quanser educational platforms, and the two companies started the collaboration to fully commercialize and launch the AELabs.

Quanser, with over 25 years of experience in design and development of teaching solutions for engineering labs, immediately noticed the strong potential synergy of the AELabs solution and the National Instruments ELVIS platform. Under the guidance of Quanser engineers, the Illuster team redesigned the AELabs control board to take full advantage of the NI ELVIS. NI ELVIS platform allowed for the integration of instrumentation, function generation, data acquisition and other core tools. The resulting AELabs for NI ELVIS offers a simpler, more compact, and portable solution for any engineering lab. Quanser helped to refine the industrial design of the boards to ensure resilience capable of withstanding even the most enthusiastic users. Keeping the original course materials closely tied to the Sedra and Smith’s Microelectronic Circuits textbook, the Quanser team also helped to align them with modern approaches to courseware design. With these enhancements, the lab experience can be now better integrated into the course, tying the theoretical knowledge to the hands-on practical learning through a set of motivating experiments.

To find out more about the research projects at the University of Rijeka, visit http://mbsdynamics-ukf.gradri.uniri.hr/ and http://canfas.gradri.uniri.hr/

Nenad Bićanić (†2016) was a Professor of Mechanics at the Faculty of Civil Engineering, University of Rijeka and Emeritus Professor at the University of Glasgow. During his career, he was also affiliated with universities in Zagreb, Croatia, Swansea, UK, and Boulder, USA. His research interests were non-linear computational mechanics and finite element method with particular emphasis on structural dynamics and modeling softening materials. He published over fifty papers in high-impact journals, supervised nearly thirty doctoral students and ran over a dozen research projects in different countries. At the University of Rijeka, he established and equipped a structural dynamics lab for experimental testing of discontinuous mechanical systems and multiple-support earthquake-induced excitation.

Gordan Jelenić is a Professor of Mechanics at the Faculty of Civil Engineering, University of Rijeka (gordan.jelenic@uniri.hr). He held previous positions at the Faculty of Civil and Geodetic Engineering, University of Ljubljana, Slovenia, and the Department of Aeronautics, Imperial College London, UK. His main research interests lie in the areas of non-linear computational mechanics, finite element method, and numerical integration in dynamics. He has published over thirty papers in high-impact journals, supervised seven doctoral students and run over half a dozen research projects in Great Britain and in Croatia. With two of his current doctoral students, he is currently working in the area experimental dynamics.

Nina Čeh is a Ph.D. candidate and teaching assistant in the field of mechanics at the Faculty of Civil Engineering, University of Rijeka (nina.ceh@uniri.hr). Her main research interests are non-linear computational mechanics, experimental dynamics of continuous and discrete structures and contact mechanics. Her focus is on experimental research of dynamic response of small model structures with the use of shaking platforms and optical measuring system.