Search

Study of the mechanical properties of glued laminated beech beams is presented in this doctoral thesis. The high strength and stiffness properties of beech wood require that special attention be paid to the tensile strength of different finger joint profiles with different adhesives used for longitudinal gluing of beech boards. The results of experimental testing of finger joints in tension have shown that longer and slimmer finger joints are required to increase the load bearing capacity of finger joints. The influence of the type of adhesive used decreases with the length of the finger joints, so that existing structural adhesives are suitable for use with glued laminated beech beams. Fourteen glued laminated beech beams were produced. The beams were then tested and the strength and stiffness properties were measured. In addition to flexural strength and flexural elastic modulus, shear modulus and contact stiffness between layers were measured. Analytical and numerical models for the analysis of glued laminated beech beams were established. The models are based on Reissner’s planar beam theory and take into account the slip between laminations and the influence of finger joints. The analytical solution for Euler-Bernoulli and Timoshenko-Ehrenfest glued laminated beech beams is presented. The numerical model can be used for modelling glued laminated beams with an arbitrary number of laminations and an arbitrary number of finger joints. The model has been validated with the experimental results and the results agree well. Stochastic analysis of glued laminated beech beams is also performed using analytical and numerical model and the results showed a good agreement with the experimental results and that numerical model can simulate the mechanical behaviour of the glued laminated beech beams. V doktorski disertaciji je predstavljena analiza mehanskih lastnosti lameliranih lepljenih nosilcev iz bukovega lesa. Zaradi visokih mehanskih lastnosti bukovega lesa je potrebno posebno pozornost posvetiti zobatim spojem ter njihovi natezni nosilnosti. Pri tem sta geometrija zobatih spojev in lepilo, ki se uporablja za dolžinsko spajanje bukovih desk, bistvenega pomena. Rezultati eksperimentalnih testov so pokazali, da z daljšimi in ožjimi zobatimi spoji nosilnost zobatega spoja narašča pri tem pa se vpliv lepila zmanjšuje. Obstoječa konstrukcijska lepila so se izkazala kot ustrezna za uporabo v lepljenih lameliranih nosilcih iz bukovega lesa. Štirinajst lameliranih lepljenih nosilcev iz bukovega lesa je bilo testiranih v upogibu. Poleg upogibne trdnosti in upogibnega elastičnega modula lepljenih lameliranih nosilcev iz bukovega lesa, smo izmerili tudi strižni modul nosilcev in togost stika med lamelami. Za analizo lameliranih lepljenih nosilcev iz bukovega lesa sta pripravljena analitični in numerični model. Oba temeljita na Reissnerjevi teoriji ravninskih nosilcev. Pri analizi je upoštevan zdrs med sloji in vpliv zobatih spojev. Analitična rešitev je prikazana za Euler-Bernoullijev in Timoshenko-Ehrenfestov nosilec. Z numeričnim modelom je mogoče modelirati lamelirane lepljene nosilce s poljubnim številom lamel ter poljubnim številom zobatih spojev. Model je validiran z rezultati eksperimentalnih testov in pokazalo se je dobro ujemanje med izracunanimi in izmerjenimi vrednostmi. Za ovrednotenje rezultatov je bila opravljena tudi stohastična analiza, tako z analitičnim, kot tudi numeričnim modelom lameliranega lepljenega nosilca.

Z optimalnim razvrščanjem konstrukcijskega lesa pravokotnega prečnega prereza lahko element uvrstimo v najboljši možni trdnostni razred v izbrani kombinaciji. V magistrskem delu smo primerjali različne postopke optimalnega razvrščanja in sicer po trenutno veljavnem standardu, po neparametričnem in parametričnem postopku. Slednja sta oba predstavljena v predstandardu. Bistvene razlike med različnimi postopki so predvsem v računu karakteristične vrednosti vzorca. Neparametrični pristop, ravno tako kot standard, pri računu upošteva poenostavitve, parametrični pristop pa temelji na teoretičnih enačbah. Za osem različnih kombinacij trdnostnih razredov smo primerjali rezultate optimalnega razvrščanja po različnih postopkih. S simulacijami v programu Mathematica smo preverili ustreznost enačb za račun karakterističnih vrednosti, s poudarkom na enačbah, ki so podane za neparametrični pristop v predstandardu. Enačba za korekcijski faktor se, glede na predpostavke predstandarda, ni izkazala za primerno, zato smo poiskali novo enačbo, s katero dobimo primernejše rezultate. Predstavljena je nova naprava za strojno razvrščanje lesa in nastavitve zanjo. Naprava še ni certificirana, zato njena uporaba za dodeljevanje trdnostnih razredov in uradno označevanje še ni dovoljena. Nastavitve so bile izpeljane na osnovi zahtev za začetne nastavitve, ki so podane v standardu. Za indikativno količino je najbolj primeren dinamični elastični modul, saj je v najboljši povezanosti s trdnostnimi karakteristikami preizkušancev. Preverili smo tudi možnost razvrščanja lesa s poenostavljenim postopkom, brez merjenja gostote. S frekvenčnimi in cenovnimi matrikami smo za izbrane kombinacije trdnostnih razredov ocenili uspešnost naprave in preverili ostale zahteve, ki jih mora naprava izpolnjevati za certifikacijo in uvrstitev na seznam naprav za strojno razvrščanje lesa, ki je podan v standardu. Izkazalo se je, da s predpostavljeno konstantno gostoto preizkušanca dobimo precej slabše rezultate. Z merjenjem gostote pa naprava daje dobre rezultate in je zato primerna za postopek certificiranja. A timber specimens highest strength grade for a selected strength grade combination can be determined by optimal grading. This can be achieved by different approaches that are represented in the current standard and pre-standard. The pre-standard defines a parametric and non-parametric calculation of characteristic properties from test results, which is the essential difference between various approaches. In both the non-parametric approach and the standard procedure, some simplifications and adjustment factors are used while in the parametric calculation only equations from a theoretical background are used. Procedures are derived for eight different strength grade combination and for all results comparisons are made. With the Mathematica software we used simulations to test given equations for the calculation of characteristic properties. Special consideration was taken on the non-parametrical calculation. The correction factor equation has not proved to be suitable, at least not for the assumptions that are made in the pre-standard. Therefore, we produced a new equation, which gives better results. A new strength grading machine is under development, which is not yet standardized and cannot formally be used for strength grading and CE marking yet. The requirements for the derivation of machine settings for initial type testing were met and we established new settings. We tried different properties as the indicative property. Due to its good correlation to specimens strength properties, dynamic modulus of elasticity was chosen. Grading procedure could be simplified by avoiding density measuring. Using size matrix and global cost matrix, we evaluated grading machine performance for data with measured density and data with constant density for each of the strength grade combination. The results using the measured density were significantly better than those with a constant density. The strength grading machine gave good results and is therefore suitable for the certification process.