23 research outputs found
Structural assessment and lateral–torsional buckling design of glass beams restrained by continuous sealant joints
No full textGlass is largely used in practice as a structural material, e.g. as beam and plate elements able to carry loads. Their structural interaction is often provided by mechanical connections, although recent trends are moving toward the minimization of metal components and the primary involvement of adhesives or silicone structural joints working as partially rigid continuous restraints.
In this work, the lateral–torsional buckling (LTB) behavior of glass beams laterally restrained by continuous silicone joints is assessed. Based on earlier contributions of literature and extended parametric Finite-Element (FE) numerical investigations, closed-form solutions are suggested for the estimation of their Euler’s critical buckling moment under various loading conditions. Finally, by means of more detailed incremental nonlinear analyses, their global LTB response is also investigated, to assess their sensitivity to initial geometrical imperfections as well as their prevalent LTB failure mechanism. In conclusion, a generalized buckling design curve able to account for the structural contribution provided by structural silicone joints is proposed for a rational and conservative LTB verification
Kipsterkte van monolithische en gelamineerde glazen liggers
Full text linkGlas verschilt van de meeste traditionele ingenieursmaterialen door zijn transparantie en zijn brosse breukgedrag, dat sterk afhankelijk is van de aanwezigheid van microscheurtjes aan het oppervlak. Sinds iets meer dan een decennium wordt glas ook aangewend in primaire draagstructuren van gebouwen, zoals glazen liggers. Tot op heden ontbreken echter normen of ontwerprichtlijnen voor dergelijke toepassingen. In dit proefschrift is theoretisch, experimenteel en numeriek onderzoek verricht naar de kipsterkte en het draagvermogen van monolithische en gelamineerde glazen liggers. Op basis van de lineair-elastische breukmechanica en de waarschijnlijkheidsleer zijn in de literatuur voorspellingsmodellen ontwikkeld voor de sterkte van glas. Omdat weinig experimentele gegevens beschikbaar zijn voor in het vlak belast glas, werden bijkomende laboratoriumproeven uitgevoerd, die het mee mogelijk maakten een rekenwaarde voor de sterkte van glazen liggers voor te stellen. De geldigheid van theoretische formules voor de stabiliteit van staalprofielen werden op experimentele en numerieke wijze geverifieerd voor glazen liggers. Deze blijken de kiplast van monolithisch glazen liggers zeer goed te kunnen voorspellen. In het proefschrift werd aanvullend een methode voorgesteld waarmee de beschikbare kipformules ook voor gelamineerde liggers met twee glasplaten en een slappe tussenlaag toepasbaar worden. Glazen liggers met een relatief massieve dwarsdoorsnede bezwijken doorgaans door te grote trekspanningen aan de onderrand, hoofdzakelijk veroorzaakt door buiging om de sterke as. Liggers met een slankere dwarsdoorsnede zijn zeer gevoelig voor instabiliteitsverschijnselen. De belangrijkste invloedsfactoren daarbij zijn de dikte, de overspanning en de stijfheid van de tussenlaag. Omdat de beschikbare tussenlagen een visco-elastisch materiaalgedrag vertonen, is de stabiliteit van gelamineerde glazen liggers ook sterk afhankelijk van de belastingsduur en de omgevingstemperatuur. In het kader van dit werk werden vergelijkende resultaten gepresenteerd voor een klassieke PVB en het nieuwere SGP. Laatstgenoemd polymeer levert beduidend betere prestaties op stabiliteitsvlak: theoretisch is het mogelijk dat de toekomstige ontwikkeling van een zeer stijve tussenlaag laminaten kan opleveren die niet alleen veiliger zijn, maar mechanisch ook beduidend beter presteren dan een massief glazen ligger met dezelfde glasdikte. Daarnaast spelen ook initiële vormfouten, excentriciteiten en de relatieve hoogte een rol. In tegenstelling tot de overige onderzochte parameters blijkt de invloed van eigenspanningen onbelangrijk. Kipgevoelige glazen liggers bezwijken door overdreven spanningen, veroorzaakt door een combinatie van buiging om beide assen en wringing. Indien dergelijke elementen gelamineerd zijn, blijkt de stijfheid van de tussenlaag een grote invloed uit te oefenen op de spanningsverdeling in beide glasplaten. Tot slot werden twee bezwijkcriteria voorgesteld voor glazen liggers, die rekening houden met de onderzochte invloedsfactoren
Influence of corner and edge distance of adhesive point-fixings for glass structures
No full textNowadays, bolted point-fixings are widely used, which implies that the glass has to be drilled and tempered. Consequently, the glass is significantly weakened exactly at the position where peak stresses occur when forces are applied. In contrast, this disadvantage does not occur when adhesive technologies are used in which the glass is directly bonded at discrete points. From bolted point-fixings, it is well-known that the distance between the connection and on the one hand the glass corner and on the other hand the glass edge are major parameters that influence the strength and behaviour of the connection. However, for adhesive point-fixings this influence has not been extensively investigated yet. Consequently, in this paper the results of a thorough experimental study of the influence of the corner and edge distance on a full-scale glass panel supported by adhesive point-fixings is published and analysed. A numerical model is constructed and validated with the experimental results. By means of a parametric study on this numerical model the effects of several parameters, such as geometrical dimensions and material properties, are studied. The experimental and numerical results demonstrate that the corner and edge distance have a significant influence on the stress distribution and deflection of the glass panel: stresses and deformations are reduced with increasing corner and edge distance. However, from a certain distance the location of maximal stress and deflection will shift from the centre to the corner and edge causing the stresses and deformations to rise again. From the parametric study the influence of several material and geometrical parameters are investigated, and an optimal distance can be found. Increasing the glass thickness, the connection diameter, the adhesive stiffness and the adhesive thickness will have a positive effect on the occurring stresse
Weak-axis flexural buckling of cellular and castellated columns
No full textCellular and castellatedmembers are usually produced by performing cutting and rewelding operations on a hotrolled
I-section member. As illustrated in previous work, these operations will influence the residual stresses present in the members in a manner which is detrimental for the flexural buckling resistance. Up to now, this has not been considered in the limited amount of literature concerning the flexural buckling resistance of these members.
In this paper, the weak-axis flexural buckling resistance is examined, taking into account the influence of the
modified residual stress pattern and the modified geometry of cellular and castellated members. Therefore, the critical buckling load and the buckling resistance of simply supported cellular and castellated members were investigated numerically. In the numerical model, a modified residual stress patternwas introduced, based on earlier measurements. As the amount of measurements was relatively limited, the results of these simulations
should be considered as preliminary results, in attendance of a confirmation of the utilized residual stress pattern.
The results of the simulations illustrate the detrimental influence of the expected residual stress pattern modification on the buckling resistance. By comparing the results with the European buckling curves, preliminary best
fit curves could be selected. This comparison was executedwith a 2T approach, inwhich all cross-sectional properties are calculated for the 2T section at the centre of the opening
New glass design method for adhesive point-fixing applications
No full textStress concentrations in glass plates connected with adhesive point-fixings are much lower than when connected with traditional bolted point-fixings. For the investigation of adhesive point-fixings a complex finite element model is usually required. Based on the SLG-method for bolted point-fixings, a design method is developed for adhesive point-fixings to simplify the finite element analysis. This is done by separating the glass plate into a local and global component. The local component represents only a part of the glass plate with one adhesive point-fixing and is modelled with more complex finite element volume elements. The global component represents the entirety and is modelled with less complex shell elements. The total stress distribution is obtained by superposition of the results of the local and global component. This paper describes how this method is developed and validated. The local numerical model is calibrated based on strain measurements obtained from experiments. The results show that the SLG-method can be suitable for adhesive point-fixings as well, which enables a simplified and faster design
Development of composite glass beams: a review
No full textIn architecture, there is a growing trend to include more transparency in structures. To increase the transparency, structural elements such as columns, beams, floors and roofs can be built in glass. However, glass is a brittle material and weak in tension, making it less evident to function as a key structural element. To provide robust and safe elements, researchers have tried to develop hybrid glass elements in which glass is combined with other materials. For the case of structural beams, composite glass beams were developed. These beams have typically T- or I-sections, in which the web is a glass laminate and the flanges are composed of another material. Both entities are then put together by using an adhesive or a bolted connection system. Also other concepts exist which have a layered section or in which the materials are combined to make trusses. In this paper, all developments and experimental investigations of this kind of glass beams are summarised using a classification based on the combined material. The concepts are evaluated considering load–displacement diagrams from bending tests. Most of the concepts described illustrate beneficial failure behaviour when subjected to bending, meaning significant post-breakage strength and ductility. Several concepts even possess a post-breakage strength that is larger than the initial glass failure load, while demonstrating extensive deformation capacity. With respect to practical applications, the latter concepts can be considered as very suitable as they generally provide significant robustness
Performance of statically indeterminate reinforced glass beams : experimental comparison with determinate systems and effect of a discontinuous glass section
No full textA well-performing concept is the reinforced glass beam, developed in analogy to reinforced concrete. This concept proved its feasibility both in statically determinate and indeterminate systems, for a variety of parameters. However, the advantages of indeterminacy over determinacy in terms of structural performance and safety are not evaluated for this concept. This paper presents experimental results of similar statically determinate and indeterminate bending tests (i.e. with the same free span length and load point positions) on stainless steel reinforced laminated glass beams, made of annealed float glass (ANG). The interlayer is used to bond the reinforcement. With these results, a comparison is made to quantify the structural efficiency of the latter system, at two test temperatures and for two reinforcement percentages.
In addition, statically indeterminate bending test results on ‘discontinuous’ glass beams are presented and compared to those of normal beams to evaluate the effect of local reduced beam stiffness on the system’s load-carrying behaviour. The first investigation led to the conclusion that a 25% increase in ultimate capacity can be reached when applying a statically indeterminate system, provided that the reinforcement-to-glass bond is thoroughly designed for the specific conditions (enabling the beam to form all necessary plastic hinges). Furthermore, when the latter condition is fulfilled, a locally reduced beam stiffness has a limited effect on the overall load-carrying behaviour of the statically indeterminate
beam system
Structural contribution of glass in Saint-Hubertus Galleries
No full textMany of the glass-covered iron and steel frames from the nineteenth and twentieth centuries now require renovation. During rehabilitation, the question arises of how to preserve built heritage while fulfilling modern standards on safety and structural integrity. In a traditional recalculation, the glass panels are considered to be dead load on the iron frame, leading to the conclusion that the original frame requires strengthening. This paper presents a calculation that takes the contribution of the glass plates into account while assessing historical glass coverings. The overall
structural behaviour of the iron and glass roof of the Saint-Hubertus Galleries was simulated and a parameter study
was performed using a finite-element software package. The original structure comprises a wrought-iron frame clad
with single glass panes connected to the glazing bars with traditional linseed-oil putty. The behaviour of the original
structure, the influence of the application of modern adhesives and sealants, and the replacement of single with
laminated glass plates is investigated. Although building such models is time consuming, including glass panes in the model leads to lower stresses, deflections and a better buckling behaviour, even if connected with linseed-oil putty
Optimal stiffening configuration for locally supported cylindrical silos: engaged columns
No full textCylindrical steel silos are often supported by discrete supports or columns to be able to provide a hopper and to facilitate emptying operations beneath the cylindrical barrel. The simplest mean of support for a light silo is by the use of engaged columns, without the use of unnecessarily expensive ring stiffeners. Such engaged columns gradually introduce the support load into the silo wall by shear, spreading the stresses in circumferential direction. In general, the highest axial compressive stress concentrations can be found in the shell wall in the vicinity of the top of the engaged column, resulting in failure due to excessive yielding and/or local instability. The study aims to identify the optimal combination of dimensions of an engaged column (i.e. the height, the widths in circumferential and radial direction and the thickness) to obtain a failure load as high as possible with as little material in the column as possible. An important condition is the requirement that the columns must withstand a higher load than the silo wall itself. In other words, failure should occur in the vicinity of the terminations of the columns (and not in the column itself). All results and conclusions are based on numerical finite element analyses
