Ref LXI_MSCA_11 - Postdoc candidate for a MSCA PF in Engineering or Physics of Damage Detection in Concrete, Structural Health Monitoring, Modal Analysis and Finite Element Modelling

The Department of Engineering (DoE) at Universityof Luxembourg is interested in hosting a postdoctoral fellow within the Marie Skłodowska-Curie Postdoctoral Fellowship programme.

Project idea

Structural Health Monitoring of bridges for damage detection is an important task for responsible institutions worldwide. If the damage is not correctly assessed, this may lead to catastrophic failure on the one hand or to unnecessary demolishment and reconstruction on the other. Above all, the damage is not always visible from outside, e.g. when special coatings are used, surfaces are concealed or difficult to access. The cracking may also appear late, i.e. close to collapse. This particular dangerous situation might happen in cases with high pretension and a low amount of passive reinforcement. Typically, old bridges from the early 1960s are known to be prone to this phenomenon. This is why those are today often recalculated to new standards. However, those bridges would be an interesting test-object for new methods and new damage indicators that are already sensitive to micro-cracking as detailed below. 

Therefore, a lot of research was done in the last decades worldwide and also at the University of Luxembourg (UL) to identify other, new damage indicators for the sake of amending traditional visual inspection by humans. Microcracking and cracking of concrete lead to local reduction of stiffness. Therefore, changes in the stiffness distribution are of high interest and may be used for damage detection, localisation and quantification. The stiffness distribution can mathematically be described by the stiffness matrix [K], which can be numerically approximated for instance with the finite element method or additionally experimentally estimated. Especially the experimental approach leads normally at first to the flexibility matrix [F], which is only the inverse of the stiffness matrix [K], i.e. [F]= [K]-1. The final inversion is important for damage detection, but is often not directly possible, as measuring errors and other limitations cannot be avoided [1], [2]. Hence, a common approach is to update Finite-Element (FE) models to have the highest agreement between the simulations and the measured characteristics that depend on the stiffness matrix such as the static bending line under known test-load or dynamic characteristics like eigenfrequencies and mode-shapes. An FE-model is first fitted to a reference state by minimising the differences between measurement and simulation. When damage with stiffness-loss appears, and one observes changes in the measurements, a second model updating process and subsequent comparison of the reference model and the current model then reveals where and how much the stiffness was reduced, i.e. damage detection, localisation and quantification becomes possible. Though this idea is quite simple, there are pitfalls in practice: 

• The stiffness of bridges depends more or less on the outdoor temperature, as e.g. YOUNGs modulus of asphalt, bearing pads, sub-soil can be temperature-dependent. The temperature effects on the measured quantities can be in the same order of magnitude as severe damage. This temperature dependency characteristic is an individual property of this bridge and can only be a-posteriori measured, i.e. once the bridge is built [1], [3]. However, efficient algorithms for temperature compensation have been developed [4], [5]. 
• Noise and limited measurement-precision truncate the accuracy of the measured characteristics [1], [6-8]. Special mobile excitation devices were developed for dynamic measurement [9-11] and high static loading without overloading is favourable [1], [6], [7]. 
• We included additionally sagging under the bridge weight as a very efficient damage indicator [8]. Due to microcracking and cracking, reinforced concrete bridges behave non-linear with not always fully reversible deformation. For centuries sagging of bridges is observed by civil engineers, which indeed is an integral parameter for non-reversible deformation within the bridge under the assumption that the abutments are fixed. Hence sagging below a zero line in the healthy reference state could be incorporated as an interesting measurement characteristic that leads to very good results using the already detailed up-dating procedure.  
• Numerical problems have to be overcome to minimise the objective function, i.e. the difference between measured and simulated quantities (static bending line, modal parameters, sagging). To avoid non-physical solutions, special clustered finite-element models have been developed and damage functions were introduced to minimise the number of adjustable parameters, i.e. Youngs-modulus of finite elements. Schommer [8] achieved very good results for a prestressed concrete beam, which was part of the former Grevenmacher Bridge (LU). This part (46 m long, 120 t) was transported to a nearby harbour after demolishment of this bridge, where it was extensively examined and then subsequently destructively tested. Artificial damage could be introduced and then detected with different static and dynamic methods. Especially the scenarios with small damage were interesting, i.e. before large visible cracks appeared. In this case, it was possible to identify, localise and quantify the stiffness loss even in case of small invisible damage. Sagging worked best, but also eigenfrequencies identified damage quickly and lead to the correct location, while eigenforms were less efficient.   

The challenge is now to extend the used model-updating methodology to non-beam-shaped bridges, i.e. bridge spans with a plate or shell-type form. A real bridge will be selected in cooperation with an external partner either in Luxembourg or Germany. It will be modelled with finite elements and static and dynamic characteristics measurements will be performed, to update the FE-model for this reference scenario. The one-dimensional Gaussian bell shape damage function with only three free parameters (location, width and peak-value) has to be extended to a second horizontal dimension. There are different possibilities for this extension, depending on the bridge type and its design. The already mentioned former Grevenmacher-Bridge, like many other bridges, was designed as a single field girder bridge, where every field consisted of several parallel I-shaped concrete beams. In that case, the used damage function could simply be repeated for any beam, knowing that there is also a transversal coupling of the beams. For real plate structures, a two-dimensional Bell curve could be defined by adding a fourth free parameter for the damage location in the horizontal direction. Generally, the damage estimation methods could be classified into two groups, namely signal-based and model-based methods. Damage is defined based on damage indices in the signal-based methods. While these methods could be successful in localising the damage, they are not appropriate for determining the damage severity. On the other hand, model-based methods could be effective for determining the severity of damage as well as its location [12], [13]. Researchers have been using different damage functions, which could also be tested [14]. Once the damage-function is defined, the effect of different levels of damage could be numerically analysed. Furthermore, other norms like the Gaussian norm could be checked numerically. Whether or not these simulations can be confirmed in real set-ups, can only be tested in a subsequent project, once a suitable bridge is found, that is devoted to demolishment. In this case, artificial damage can be introduced and its effect studied. This final set is probably not possible within this demanded Marie Skłodowska-Curie Actions Postdoctoral Fellowship. 

Offer Requirements

The ideal post-doc candidate needs an engineering background in mechanical-, civil- or even electrical-engineering, applied mathematics or physics. A ‘very good’ or ‘excellent’ PhD degree in one of the previously mentioned domains is expected. The candidate is willing to step into structural health monitoring of concrete bridges and finite element modelling and numerical optimisation. Though the project is before all a numerical study, experimental work is needed to measure and subsequently model the reference state of a shell-type bridge. Team-work with a few other colleagues at UL and with external partners (Administration des ponts et chaussées (LU), company LUCKS Technologies (DE)) is essential.

Institute description

The University of Luxembourg is composed of three faculties and several specialized centres for distinct research. The Faculty of Science Technology and Medicine (FSTM) regroups traditional university teaching and research missions and hosts among others the Department of Engineering (DoE), which combines mechanical, civil, electrical and computational engineering. The research plan of the department targets landmarking Luxembourg on the scientific map in engineering sciences for teaching and research to support all engineering sectors in Luxembourg. Research and teaching within the unit are coordinated by 21 professors/associate professors, 10 lab engineers and administrative staff. There are actually more than 60 PhD students enrolled in the engineering doctoral schools. In the last 10 years four research projects with doctoral diploma were successfully completed in the group of Prof. Stefan Maas. Those focussed on lower limb and its diverse implant. The team’s research work is published in scientific journals of the field and international conferences and may be found on Search results.

About Marie Skłodowska-Curie Actions Postdoctoral Fellowship

The MSCA PF are postdoctoral fellowships, financed through the European Commission for 2 year positions, covering salary and research costs of the researchers, who can come from anywhere in the world. The evaluation process of these programmes is more about the project and the career development of the candidate, than his/her publication record. The organizations in Luxembourg that want to host such talented researchers include large public research organizations as well as private companies that offer interesting job perspectives after completion of the fellowship.

Minimal eligibility criteria

• Individual: applicant applies together with host institute
• PhD at the deadline (PhD date cannot be more than 8 years ago)
• Not working or living in Luxembourg for longer than 12 months before the call deadline
• For rules see here or contact your NCP

How to apply

Please fill out the application form until May 15^th 2021 – please note that we cannot accept applications without the application form. We need your CV with a list of publications and if relevant a short project proposal (maximum 3 pages) as well as the offer reference you are interested in. You can also send several fine-tuned CVs if you are interested in multiple positions – please fill out a new application form in that case. Make sure you address the text and requirements mentioned by the respective hosts.

Please note that applications are made through Luxinnovation, your contact point for Horizon Europe in Luxembourg, before being handed over to the host company.

Don’t hesitate to contact us, if you have any questions.

Luxinnovation, the Luxembourgish Innovation agency and National Contact Point for MSCA, will support both you and your future host institute during your application phase. For all applicants coming to Luxembourg we offer guidance documents, webinars, training in proposal writing and we will review your proposals if you wish so (depending on our availability). In 2019, MSCA IF proposals from Luxembourg had an average success rate of 28.5%, and proposals reviewed by our service 42%.

More information can be found on our website.

See other similar positions here.

KEYWORDS

• Damage Detection in Concrete Bridges
• Structural Health Monitoring
• Modal Analysis
• Finite-Element Modelling
• Model-Updating and Optimisation
• Damage Functions
• Influence Lines
• Flexibility Matrix

RESEARCH IN LUXEMBOURG

• Luxembourg is 2: both country and city ! Great places 2 live and work !
• Let's make it happen
• Research Luxembourg