Search

The small-scale physical model platform was developed as part of a four-year research project, "Physical modeling of landslide remediation constructions' behaviour under static and seismic actions, " to investigate the behaviour of slopes built of different materials with and without remedial measures under the influence of rainfall and earthquake. As in real landslides, the observation of a physical model consisted of a geotechnical and a geodetic monitoring system. In this paper, the behaviour of clay-like slopes during 1g artificial rainfall, with and without remedial measures will be described. Complex remedial measures included deep trench drains which reduced pore water pressure, and pile wall with head beam made with 3D printer, which increased shear resistance at the sliding surface. At the same rainfall intensities, the clay-like slope collapsed, while the slope supported with a pile wall at the slope foot remained stable. The results of the photogrammetric equipment and the complex sensor network measurements within a slope are presented and discussed, helping us to understand the different behaviour of slopes with and without remedial measures.

The paper presents an analysis of stresses and strains in both finite and infinite rectangular homogeneous, isotropic elastic plates with a circular hole in the middle loaded by a uniform tensile traction.

Small-scale physical models of landslides triggered by rainfall and seismic conditions provide a good insight into the initiation and progression of full-scale landslides in nature. In order to track and document the displacements on the surface of the small-scale model, a digital image correlation-based optical measuring system with high-speed cameras is used here. Each model is prepared for the optical measurements by adding specially chosen marker points (pins) that are monitored by a pair of high-speed cameras during each experiment. An additional set of non-high- speed cameras with higher resolution is used to monitor the deformation field on a selected smaller part of the model. This enables to obtain the 3D displacements and velocities of each marker point in order to detect any movement or crack opening on the surface both visually and accurately from the optical measurement results. The described system and established measurement procedure are advantageous as they provide the 3D displacement and velocity data for a large number of points on the surface with less equipment than conventional contact measurement methods. The collected data, in combination with other monitoring sensors, allow the observation of landslide initiation and the analysis of landslide evolution in all parts of the model slope during the sliding process. In this paper we present the measurement procedure and the results obtained optically in selected small-scale experiments.

Earthquake loading might represent the worst loading condition the structure will experience in its design life. If not accounted for it properly, the loading might be devastating for the structure and people using it, as recently witnessed during the devastating earthquakes in north-western Croatia. However, overestimating the loading might lead to very redundant structures, an unnecessary waste of resources, with negative implications to sustainability. This work presents results obtained by a numerical procedure (based on finite element method with 3D beam Euler-Bernoulli and Timoshenko elements, Rayleigh damping theory and Newmark's method of numerical integration) as well as an experimental program on a simply supported beam model (conducted by means of two biaxial seismic platforms and optical measuring system).

This paper presents original experimental research on soil-structure interaction conducted by authors. Small scale steel frame model was tested using shaking table. The model comprised two columns, a foundation beam representing strip foundations and a rigid beam connecting the top of the columns representing the superstructure mass. Model was tested as being fixed into the shaking platform and as placed on sand layer embedded into a container. Displacement time history was recorded using optical measurement system and compared for every tested case. Sand used in experimental research was local Drava sand. Experimental research is repeated for the case of dry and saturated sand. Results for all tested cases are shown and commented.

Physical modelling of landslides by analysing the behaviour of small-scale landslide models subjected to artificial rainfall can be divided into modelling under 1g conditions and under increased acceleration (n times gravity) in a centrifuge. In this paper, the landslide initiation, progression, and deposition caused by artificial rainfall in three small-scale models built with sand or sand-kaolin mixtures at the same slope angle will be described. The evolution of landslides is monitored by observation of volumetric water content, matric suction, and pore water pressure, as well as by monitoring slope deformations and failure development. Analysis of the factors affecting the landslide initiation, propagation and their relationship to the slope material, the infiltration process, and the overall resistance of the soil in a slope in terms of soil strength, effective pressure and the contribution of matric suction in the unsaturated part of the slope will be discussed. The main observations from the results of the tests carried out in relation to the initiation and development of the observed instabilities of sandy and clayey slopes are given.

Physical modelling of landslides using scaled landslide models behaviour began in 1970s in Japan on scaled natural slope models. The laboratory experiments of landslide behaviour in a scaled physical model (flume or flume test) started in 1980s and 1990s in Canada, Japan and Australia under 1g conditions. The main purpose of the landslide physical modelling in the last 25 years was research of initiation, motion and accumulation of fast flow like landslides caused by infiltration of water in a slope. In this paper the results obtained in landslide initiation test of a sandy and clayey slopes at different slope angles exposed to artificial rain by rainfall simulator are presented. The results of landslide development are monitored by observation of volumetric water content, matric suction and pore water pressure, as well as by monitoring of surface displacements and the displacements inside the sliding mass. The main observations from the results of provided tests related to initiation and development of the observed instabilities of sandy and clayey slopes are given.

In this work the Moreau-Jean θ-scheme is applied in the simulation of a problem involving a rigid block rocking on an elastic beam.

Cilj ovog rada je numerički modelirati ljuljanje bloka na grednoj konstrukciji. Blok je modeliran kao kruto tijelo, dok je za gredu uzeto da se može deformirati prema teoriji greda. Numeričko rješenje kretanja grede dobiva se metodom konačnih elemenata. S obzirom da problem uključuje i pojavu kontakta između grede i bloka, modeliranje će se bazirati na ne-glatkoj kontaktnoj dinamici.

Ljuljanje predstavlja važan doprinos dinamičkom odgovoru diskontinuiranih sustava, poput spomenika, suhozida, grafitnih blokova u AGR nuklearnim elektranama, ali i zidanih konstrukcija nakon popuštanja veziva. Ovdje je analitičkim, numeričkim i eksperimentalnim pristupom istraženo ljuljanje. Ponovno su izvedene jednadžbe kretanjaza ljuljanjejednog krutog bloka i stupca od dva kruta bloka uslijed pobude podloge. Jednadžbe su diskretizirane Newmarkovom numeričkom integracijom te riješene u geometrijski nelinearnom obliku Newton-Raphsonovom iterativnom metodom. Osmišljenje i proveden niz laboratorijskih eksperimenata slobodnogi prinudnogljuljanjaaluminijskih blokovaza cijeli niz vitkosti i nekoliko različitih veličina, čime je istražen efekt skale/veličine i efekt vitkosti nanjihovdinamički odgovor. Posebno važan mehanizamu kod ljuljanja je disipacijaenergije, kojase većim dijelom događa u sudaruizmeđu koja se ljuljaju. Numeričkim simulacijamaje pokazano kakovitkost i veličina blokova utječu na disipaciju energije idinamičkustabilnost blokova, te je isti utjecaj eksperimentalno validiran.

In this work, dynamic characterisation of blocky structures is investigated. First, rocking of blocks is investigated. By this model, the behaviour of historical monuments, dry-stones, blocks in nuclear power plants, and masonry structures after the failure of the mortar may be described. Non-linear equations of motion are derived and solved numerically. The exact time of impact is detected and the kinematic contact conditions after the impact are derived. Experimentally obtained energy loss is introduced. The algorithm is validated using a specially designed set of experiments, which also represent benchmark cases for validation of rocking. Conditions for complete or partial overturning of a dual-block stack are investigated experimentally and numerically. Second, organisation of an ordered multiple-block assembly inside a basin due to harmonic excitation is investigated. The parameters for detection of diverse modes of organisation and characterisation of undesirable events are defined from experimental observation., U ovom radu se istražuje dinamička karakterizacija blokovskih sustava. Ponašanje blokovskih sustava može opisati ponašanje povijesnih spomenika, suhozidanih konstrukcija, blokova u nuklearnim elektranama i zidanih konstrukcija nakon popuštanja veziva. Najprije se istražuje ljuljanje samo jednog bloka pod utjecajem pobude baze na kojoj se blok nalazi. Nelinearne jednadžbe kretanja koje opisuju ljuljanje jednog bloka su izvedene i riješene numerički koristeći Newmarkovu metodu za numeričku integraciju i Newton-Raphsonovu metodu za iterativno rješavanje nelinearnih jednadžbi. Detektirano je točno vrijeme sudara između tijela, što omogućuje određivanje brzine i ubrzanja sustava neposredno prije sudara. Izvedene su i brzine neposredno nakon sudara iz uvjeta da je moment količine kretanja sustava tijekom sudara održan obzirom na točku koja predstavlja novi centar rotacije. Odnos između brzine neposredno prije i neposredno nakon sudara je definiran koeficijentom restitucije. Posebno osmišljenom i provedenom serijom eksperimentalnih ispitivanja slobodnog ljuljanja, uz pomoć sustava za beskontaktno optičko mjerenje pomaka i deformacija GOM Aramis i Pontos, dobivene su stvarne vrijednosti koeficijenta restitucije. Pokazano je da tako dobivene vrijednosti potvrđuju potrebu za korištenjem modificirane formule za izračun koeficijenta restitucije. Eksperimentalno dobivene vrijednosti su uvedene u numerički model. Algoritam je eksperimentalno potvrđen pomoć posebno osmišljenog skupa pokusa, koji također predstavljaju benchmark primjere za validaciju simulacija ljuljanja: ljuljanje (i prevrtanje) bloka prilikom konstantnog ubrzanja podloge određenog trajanja (provedeno na air-track uređaju), te ljuljanje (i prevrtanje) prilikom jednog vala sinusoidalnog ubrzanja podloge (provedeno na preciznoj potresnoj platformi Quanser ST-III). Stupac koji se sastoji od dva bloka jednakih širina koji stoje jedan na drugome se prilikom ljuljanja može naći u osam različitih konfiguracija. Nelinearne jednadžbe kretanja za ljuljanje u svih osam konfiguracija su izvedene iz Lagrangevih jednadžbi. Definirani su prelasci između konfiguracija, koji se događaju ukoliko su kinematički uvjeti za aktivaciju ’više’ konfiguracije zadovoljeni, ili ukoliko je došlo do sudara između tijela u sustavu (ili između gornjeg i donjeg bloka ili između donjeg bloka i podloge). Na temelju navedenog je napisan numerički algoritam pomoću Newmarkove metode za integraciju i Newton-Raphsonove iterativne metode, te procedure za detektiranje točnog vremena svakog sudara između tijela. Brzine u sustavu nakon sudara su izvedene iz uvjeta da moment količine kretanja gornjeg bloka te moment količine kretanja cijelog stupca obzirom na točke koje predstavljaju centre rotacije nakon sudara moraju biti održani pri sudaru. Algoritam je eksperimentalno potvrđen serijom ispitivanja ljuljanja stupca od dva bloka prilikom jednog vala sinusoidalnog ubrzanja podloge. Uvjeti za potpuno (oba bloka) ili djelomično (samo gornji blok) prevrtanje stupca su istraženi numerički i eksperimentalno. Stupac koji se sastoji od tri bloka je eksperimentalno ispitan na posebno osmišljenoj platformi za dvostruki impuls ROORI1 na Sveučilištu u Oxfordu. Ispitani su uvjeti (posebice vrijeme između dva impulsa pobude baze) pod kojima dolazi do prevrtanja samo najgornjeg bloka, dva gornja bloka i cijelog stupca. Na kraju, istražena je organizacija uređene skupine od više blokova unutar posude prilikom horizontalne harmonijske pobude. Parametri za detekciju različitih oblika organizacije, periodično ponašanje te karakterizaciju nepoželjnih događaja, poput položaja težišta sustava te momenta inercije sustava s obzirom na težište, su definirani na temelju eksperimentalnih promatranja.

Rocking stability of a rigid prismatic block standing on a rigid base subject to a single-sine or a single-cosine wave acceleration function is examined. The stability for various slendernesses and sizes is assessed numerically, where an improved coefficient capable of estimating the size effect is taken into account. A number of relevant cases are validated experimentally, with a specially designed set of rocking benchmark tests on a shaking table system.

U seizmički aktivim područjima dugačke konstrukcije s velikom udaljenosti između oslonaca ili osloncima na podlogama različitih geoloških osobina mogu biti pod utjecajem različite seizmičke pobude na svaki od oslonaca. Posljedica toga može biti oštećenje ili otkazivanje sklopa konstukcije uslijed relativnih pomaka oslonaca. U ovom radu je analiziran dinamički odziv jednostavnog grednog nosača na nejednoliku pobudu oslonaca. Provedeno je eksperimentalno ispitivanje dinamičkog odziva modela slobodno oslonjene grede s dodatnim koncentriranim masama, gdje je pobuda nanesena koristeći sustav od dvije potresne platforme, a odziv mjeren pomoću beskontaktnog 3D mjernog sustava. Poznati analitički pristup za analizu ovakvog problema (modalna analiza), koji u obzir uzima samo linearne efekte ponašanja, primijenjen je i uspoređen s eksperimentalno dobivenim rezultatima. Određene su amplitude seizmičke pobude unutar kojih je odgovor grede linearan. Dani su eksperimentalno dobiveni rezultati pomaka stupnjeva slobode grede u vremenu prilikom harmonijske jednolike i nejednolike pobude te prilikom simulacije Northridge potresnog zapisa koji započinje istovremeno na oba oslonca ili s faznim pomakom na jednom od oslonaca. Dani rezultati predstavljaju benchmark probleme za validaciju budućih numeričkih pristupa ovom problemu.

Free rocking of a prismatic rigid block is analysed and a numerical procedure based on precise contact detection is developed and tested. In addition, a comprehensive parametric study of free rocking of blocks of different slenderness and size is conducted experimentally. Post-impact behaviour and energy loss in relation to existing restitution models (Housner’s classical model [1] and the improved model by Kalliontzis et al. [2]) is investigated in detail. Based on the numerically and experimentally obtained results, the use of the improved model is strongly encouraged, and a procedure to estimate the unknown parameter in the improved formula is suggested. [1] G. W. Housner, “The behavior of inverted pendulum structures during earthquakes, ” Bull. Seismol. Soc. Am., vol. 53, no. 2, pp. 403–417, 1963. [2] D. Kalliontzis, S. Sritharan, and A. Schultz, “Improved Coefficient of Restitution Estimation for Free Rocking Members, ” J. Struct. Eng., vol. 142, no. 12, 2016.

Rocking of prismatic rigid blocks is analysed and a numerical procedure based on precise contact detection is developed and tested. First, a parametric study of free rocking of blocks of different slenderness and size is conducted experimentally. Energy loss in relation to existing restitution models (Housner’s classical model [1] and the improved model by Kalliontzis et al. [2]) is investigated in detail. Based on the numerically and experimentally obtained results, the use of the improved model is strongly encouraged, and a procedure to estimate the unknown parameter in the improved formula is suggested. Secondly, stability full time response due to a single sine- or cosine-wave acceleration excitation is obtained numerically, where the improved restitution model [2] with the suggested estimated parameter is used. Finally, a chosen set of numerical cases is validated experimentally using a specifically designed set of experiments involving shaking tables and optical measurement equipment. REFERENCES [1] G. W. Housner, “The behavior of inverted pendulum structures during earthquakes, ” Bull. Seismol. Soc. Am., vol. 53, no. 2, pp. 403–417, 1963. [2] D. Kalliontzis, S. Sritharan, and A. Schultz, “Improved Coefficient of Restitution Estimation for Free Rocking Members, ” J. Struct. Eng., vol. 142, no. 12, 2016.

Predmet istraživanja jest dinamička analiza odziva grednog mosta na prostorno promjenjivu potresnu pobudu, odnosno pobudu uslijed različitog pomicanja oslonaca. Budući da je Građevinski fakultet u Rijeci jedna od nekolicine ustanova u svijetu koje posjeduju čak dva stola za potresnu simulaciju, sve analitički dobivene rezultate bit će moguće potkrijepiti i eksperimentalnim dijelom istraživanja. Rad obuhvaća formulaciju jednadžbi problema i postupke njihova rješavanja, a u sklopu eksperimenta ispitane su slobodne oscilacije slobodno oslonjene grede i konzole s koncentriranim masama te prisilne oscilacije uslijed pomaka oslonaca. Naposljetku je dana usporedba analitički i empirijski dobivenih rezultata.

Kruti blok ili stupac krutih blokova koji se ljulja uslijed pomicanja baze na kojoj se nalazi (bez proklizavanja i odskakivanja) je opisan skupom od osam jednadžbi kretanja za osam različitih konfiguracija u kojima se može naći. Prelazak iz jedne konfiguracije u drugu se može dogoditi zbog zadovoljavanja uvjeta odnosa opterećenja baze i dinamičkih veličina koje opisuju kretanje blokova ili zbog sudara među blokovima. Kod numeričkog rješavanja jednadžbi kretanja je potrebno detektirati točan trenutak sudara, prilikom čega početne vrijednosti dinamičkih veličina u novoj konfiguraciji, u idealnom slučaju, moraju zadovoljavati jednadžbe balansa energije i balansa momenta količine kretanja. Za analizu ovoga problema napisan je numerički algoritam u programskom jeziku Matlab. S ciljem validacije algoritma, osmišljen je niz kontroliranih laboratorijskih eksperimenata ljuljanja jednog bloka i stupca od dva bloka uslijed konstantnog ubrzanja podloge određenog trajanja. Opisan je mehanizam kojim su kontrolirani stupnjevi slobode blokova i postupak nanošenja konstantnog ubrzanja podloge. Osim usporedbe eksperimentalno i numerički dobivenih rezultata, analizirana je ovisnost oblika prevrtanja (djelomično ili globalno prevrtanje) o parametrima ubrzanja baze kod dva bloka.

A prismatic block is analysed for overturning when subject to a constant ground acceleration of prescribed magnitude and duration. The analytical overturning condition and a special numerical procedure are developed and assessed against experiments. Generalisation of the approach to rocking of a dual-block stack is outlined.

The subject of the present study is dynamic response of beam bridges under spatially variable ground movement (the so-called multi-support excitation). The phenomenon of spatial variability of ground motion mostly affects larger structures with greater distance between supports, such as long bridges, dams, tunnels etc., playing an important role in their safety and durability. Such effects may result in structural failure due to excessive relative displacements between the supports. The study includes formulation of the free and forced vibration equations along with solution procedures applied to simply-supported and cantilever beam systems. The modal analysis will be used as the method to calculate the dynamic response (displacements, velocities and accelerations of the nodal points as well as the internal forces) of such systems to support excitation. Since the Faculty of Civil Engineering in Rijeka possesses two shake tables, all the obtained results will be compared with those provided experimentally. For the experimental part, the 3D non-contact optical displacement and deformation measurement system will be used for data management and post-processing. Computational analysis of structures, especially large ones, subject to spatial variability of ground motion in seismic areas is still relatively unexplored and further research supported by controlled laboratory testing is highly recommended for future construction planning and structural design.

In this work structures that consist of one or two rigid rectangular blocks (one on top of the other) which are not connected between themselves or to the ground, and undergo only rocking motion under dynamic excitation are investigated. Analytical description of the motion for different rocking patterns is presented, transition criteria and exact time of each impact are derived, and energy and momentum balance during rocking motion are addressed. The equations of motion are discretized using Newmark’s method and solved at each time step using the Newton-Raphson procedure. Modes of overturning and pattern during rocking motion due to a constant ground acceleration function are looked into.

Stupac slobodno položenih blokova pri dovoljnoj dinamičkoj pobudi odgovara gibanjem sastavljenim od klizanja po podlozi, ljuljanja (rotacija) oko uglova i podizanja/poskakivanja pri udarima o podlogu. Složeni dinamički problem više tijela s međusobnim kontaktima se najčešće modelira nekom od metoda diskretnih elemenata, uzimajući u obzir geometriju i materijalne karakteristike blokova i karakteristike kontaktnih ploha. U radu se promatra slučaj dvostruke impulsne pobude podloge, te se istražuju uvjeti za postizanje drugih (tzv. „viših“) oblika djelomičnog prevrtanja stupca, koji se pri prevrtanju podijeli na više grupa blokova. Proveden je niz eksperimenata sa stupcima slobodno položenih blokova na kliznoj podlozi podvrgnutoj dvojnom udarnom impulsu, u kojima je variran intenzitet prvog udarnog pulsa i vremenski interval nakon kojeg slijedi povratni impuls. Prvi impuls je nanesen ispaljivanjem projektila, a intenzitet je kontroliran različitim inicijalnim pozicijama opruge. Povratni impuls je rezultat sudara klizne baze s fiksnom preprekom, a vremenski interval je kontroliran modifikacijom udaljenosti između klizne baze i prepreke. Rezultati svih eksperimenata su sistematizirani na temelju karakterizacije ishoda eksperimenta (trajne stabilnosti, djelomičnog ili globalnog prevrtanja stupca blokova). Detaljnije su istraženi uvjeti intenziteta prvog impulsa i vremenskog intervala između dva impulsa pod kojima se pojavljuju „viši“ oblici prevrtanja, odnosno djelomično prevrtanje stupca.

Rocking motion of a single rigid rectangular block on rigid base subjected to base acceleration function, with the assumption that friction between block and base is large enough to prevent sliding motion, is analyzed. In literature, such analysis has been carried out using a linearized equation of motion assuming slender geometry of the block and small rotations. The analysis herein additionally addresses the cases when the block is non-slender and can undergo large rotations. With the aim to investigate the influence of assumptions introduced into linearized analysis on the accuracy of solution, three types of analysis are carried out: (1) linearized analysis assuming slender geometry of the block, (2) quasi-nonlinear analysis, i.e. linearized analysis taking real geometry of the block into account, and (3) nonlinear analysis. Newmark0s method with average acceleration is used for numerical integration and Newton-Raphson iterative procedure is used to solve the nonlinear equation. Free and forced rocking motion of slender and non-slender rigid rectangular block with varying initial conditions are analyzed. Analytical solutions of the linearized and quasinonlinear equation of motion are compared to numerical results.

Dynamic sensitivity of a one-dimensional assembly of eight rigid blocks undergoing external harmonic vibrations is investigated. Several parameters are considered as dynamic attributes: time variation of the mass inertia of the assembly and for each block the average over all the simulation of granular temperature, relative velocity, coordination number. Numerical simulation is based on the Non Smooth Contact Dynamics (NSCD) time integration framework Solfec (http://code.google.com/p/solfec/). Sensitivity space parameters include a range of excitation frequencies and velocity amplitudes.

There is a general lack of well controlled benchmarks to validate predictive capabilities of computational simulations relevant for multiple-block structural configurations with clearances (DEM and non-smooth contact dynamics, NSCD). Closed form analytical benchmarks are largely restricted to 2D single or double rigid block stacks, typically concerned with rocking and overturning conditions due to harmonic or step base excitation. Herein both experimental and computational dynamic sensitivity study of multiple-block stacks subjected to pulse base excitation are examined. Advanced noncontact optical measuring technique (GOM Aramis and Pontos system and the associated processing software) have been applied to replace conventional measuring techniques. The NSCD simulation framework SOLFEC is adopted here, which effectively ignores the high frequency content of the contact interactions. Instead of a specified interpenetration-force relation, this paradigm employs the complementarity relation between the relative velocity and the contact impulse at an existing contact point. This velocity-impulse relation is added as an algebraic constraint to the implicitly integrated momentum balance and the ensuing nonlinear contact problem is therefore solved implicitly at every time. Series of test experiments were conducted in the Oxford Impact Engineering Laboratory on a bespoke platform for a controlled pulse base excitation. Impact is generated by a pin-ball mechanism attached to an optical bench, where the teflon guide and stopper were aligned to the impact device and also attached to the optical bench. Rubber cushions were used to control the shapes of the initial and reverse impact signal. Every experiment was recorded with Phantom or Photron video camera with frame rate of 2000 fps. Comparative SOLFEC analyses were conducted as a validation study. For the simulations the base was subjected to a constant acceleration of a finite duration until the required base velocity was achieved. Overturning modes in simulations and experiments were characterized as a function of projectile velocity (or indirectly by initial velocity of the base) and the stop gap distance. The conducted set of benchmarks for the validation of simulation paradigms for discontinuous media is believed to be valuable for researchers and code developers (NSCD, DEM, DDA), as well as for safety case engineers and industry regulators.

Experimental and computational dynamic sensitivity study of Multi-Block Stacks subjected to Pulse Base Excitation is considered. Advanced non contact optical measuring technique based on the GOM Aramis and Pontos systems, as well as the corresponding processing software (displacement history of control sensor points, with a high resolution high speed cameras) have been applied to replace conventional displacement measuring systems and accelerometers. The Non Smooth Contact Dynamics (NSCD) time integration simulation framework SOLFEC http://code.google.com/p/solfec/ is adopted here for comparative NSCD analyses, including a sensitivity study on interface characteristics, as a validation process. Series of test experiments were conducted and recorded on a bespoke platform with and without lateral constraints in the Oxford Impact Engineering Laboratory and Rijeka University Structural Dynamics Laboratory for an extensive series of controlled pulse base excitation tests of multi block stacks configurations. Impact is generated by a pin-ball mechanism with spring and a wooden projectile, attached to an optical bench. For the NSCD simulations the base was subjected to a constant acceleration over a finite time, thereby facilitating the characterisation of multi block stacks tumbling modes of failures (global or partial), as a function of stop gap distance. Creation of well documented benchmarks for the validation of simulation paradigms for discontinuous media will be extremely valuable for researchers and code developers (non smooth contact dynamics, discrete elements, discontinuous deformation analysis), as well as for safety case engineers and industry regulators.

Colloque avec actes et comité de lecture. Internationale.; International audience; Dynamic sensitivity of a one-dimensional stack of four rigid blocks with controlled initial gaps, undergoing external harmonic vibrations is investigated. Time variation of the mass inertia and the relative kinetic energy are considered as dynamic block stack attributes. Numerical simulation is based on the Non Smooth Contact Dynamics (NSCD) time integration framework Solfec (http://code.google.com/p/solfec/). Sensitivity space parameters include a range of excitation frequencies and velocity amplitudes.

U radu se promatra ponašanje ne-glatkog dinamičkog sustava – odziva krutog bloka na podlozi, uz pretpostavku da je trenje između baze bloka i podloge dovoljno veliko da je spriječeno klizanje bloka po podlozi. Na blok djeluje dinamička pobuda harmonijskog ubrzanja podloge. Numerička integracija jednadžbe kretanja je provedena Newmarkovom metodom prosječnog ubrzanja (β=1/4, γ=1/2), a gubitak energije pri sudaru baze bloka i podloge je uzet u obzir pomoću koeficijenta restitucije. Rezultati su prikazani u obliku grafova zaokreta, kutne brzine i kutnog ubrzanja u vremenu, te u faznoj ravnini. Promatrana je periodičnost dinamičkog ponašanja i simetričnost, odnosno nesimetričnost, orbita u faznoj ravnini. Analizirana je osjetljivost orbita u faznoj ravnini na male perturbacije početnih uvjeta vertikalnosti bloka. Simulacije ne-glatkog dinamičkog ponašanja krutih blokova pokazuju postojanje periodičnosti, ovisno o parametrima sustava (geometriji bloka, koeficijentu restitucije i funkciji ubrzanja podloge) te početnim uvjetima nagnutosti bloka. Uočena je nesimetričnost orbita pri određenim perturbacijama početnih uvjeta i, kao posljedica, manja amplituda funkcije ubrzanja podloge potrebna za prevrtanje bloka.

U radu su eksperimentalno određena svojstva svježeg i očvrsnulog procjednog betona. Analizirana je zavisnost tlačne čvrstoće, vlačne čvrstoće savijanjem te koeficijenta protočnosti o granulometrijskom sastavu agregata. Budući da ne postoji propisani način određivanja protjecanja vode kroz procjedni beton, ispitivanje je provedeno sukladno poznatim metodama, ali s pomoću samostalno izrađenih pomagala. Na uzorcima s diskontinuiranom granulometrijskom krivuljom, gdje udio frakcije pijeska 0-2 mm u sastavu mješavina nije prelazio 15 %, dobiveni su optimalni rezultati čvrstoće i protočnosti.