Search

Assumptions and simplifications in finite element (FE) modelling for structural vibration analysis contribute to discrepancies in numerical results. This paper presents FE model updating on components of a simplified model of an aircraft pylon. The components are tested under free-free boundary conditions. Impact testing is performed with roving accelerometers technique. Modal parameters are predicted numerically using NASTRAN SOL 103 normal modes analysis. NASTRAN SOL 200 optimisation for modal based updating method is subsequently used and successfully minimised the total error of the initial FE model from 19.17 per cent to 3.19 per cent. These satisfactory results could be confidently enhanced the quality of the FE model to be used for the further engineering analysis. [ABSTRACT FROM AUTHOR]

The laser spot weld is widely used to join automotive structural components due to the advantages of being inexpensive and more efficient. However, developing a reliable FE model of a laser spot weld for investigating a structure's modal parameters is very challenging. A repeated assessment of the welded joint model can significantly affect the efficiency of the finite element method, especially for a large complex structure that usually consists of huge numbers of degrees of freedom. This paper presents a wave-based substructuring method to speed up a modal parameters analysis of a laser spot welded structure. A reduced model is developed based on the wave-based substructuring scheme and compared with the full finite element model in computational time and accuracy. It was found that the proposed scheme was able to reduce the computational time of a modal parameters analysis significantly. The proposed scheme may provide advantages for research requiring high computational analysis. [ABSTRACT FROM AUTHOR]

Bolted joint structure is one of the complex types of the assembled structure that is applied in different sectors of industries. The benefit of the joining method using bolt and nut including low production cost, low maintenance and simplistic design of assembly. The prediction of dynamic behaviour of the bolted joins assembled structure is very challenging. The dynamic behaviour of the assembled structure such as bolted joined structure can be altered by the uncertainties of parameters such as material properties, structural geometry, and boundary conditions of the structure. The uncertainties that are inherited in the assembled structure can be analysed and evaluated using stochastic structural response by benchmarking the obtained results with the experimental data. In this study, a few bolted joint models are represented with RBE2, CBAR and CBEAM element connectors and initially constructed using finite element method (FEM). The initial response of the FE models with different type of element connectors in terms of natural frequencies and mode shapes are compared with experiment counterpart. The initial bolted joins assembled structure using CBEAM have shown the highest correlation response in terms of natural frequencies and mode shapes. The analysis using Stochastic Finite Element Method (SFEM) by utilising The Modified Extensible Lattice Sequence (MELS) algorithm is then used to predict stochastic responses from the initial FE model of the bolted joins structure. The result shows the adaptation of MELS algorithm on the analysis of the bolted joins assembled structure with CBEAM shows a better correlation to the experimental counterpart as higher number of samples applied to the structure. [ABSTRACT FROM AUTHOR]

Acquiring efficient, economic and accurate methods for identifying the material properties of heterogeneous materials have been one of the major objectives in the mechanical engineering community, thus, inverse methods may be considered as contributing to a very promising way of achieving the objective. This paper is concerned with an inverse method for identifying the material properties of a kenaf reinforced composite test model. An initial finite element model of the composite test model is developed and normal modes analysis is carried out to calculate the modal parameters of the initial finite element model. The inverse method which is developed from MSC Nastran SOL200 is used to update iteratively the material properties of the initial finite element model based on the comparison between the calculated modal parameters and those observed experimentally from experimental modal analysis (EMA). The objective is to acquire the most appropriate set of material properties that provide well match between the finite element model and the test model. It was found that the identified material properties are well matched with that of previous published papers in terms of the reduction in total errors, which are discussed in this paper. [ABSTRACT FROM AUTHOR]

Adopting dynamic substructuring methods to predict the dynamic behaviour of a combined structure is a common practice in the field of structural dynamics. This method has the capability and versatility to combine different sources of data; experimental and analytical, via the frequency based substructuring (FBS) method. This paper investigates the dynamic behaviour of a structure with doubled bolted joints using the frequency based substructuring (FBS) method. In the analysis the substructures are combined in the frequency domain, and the frequency response functions of the substructures are analytically derived and experimentally measured. However, the accuracy of the FBS method is highly dependent on experimental rotational degrees of freedom (DOF) which are always found to be very difficult to measure accurately. Therefore, a new approach - namely single-axis multipoint connection through which the need for the rotational DOF in the FBS method can be neglected entirely is presented in this study. The actual test structure used in this study is an assembled structure consisting of two substructures, a simple beam (numerical model) and an irregular plate steel (experimental model) structure. The FRFs of both substructures are combined via the FBS method. This study reveals that the use of the FBS method with the proposed approach is capable of representing the experimental frequency function response accurately even without rotational degrees of freedom data. [ABSTRACT FROM AUTHOR]

Accurate representation of bolted joints in the finite element (FE) model of a built-up structure is a crucial factor in ensuring the accuracy and reliability of the dynamic behaviour prediction of the structure. Modelling of bolted joints in industrial structures has not been widely and thoroughly studied, especially in the dynamic response of the bolted joint itself. The aim of this work is to investigate several flexibility formulae for the calculation of the stiffnesses of CBUSH elements for the modelling of the bolted joints. The flexibility formulae used in this work are attributed to Huth, Tate & Rosenfeld, Boeing, Swift and Grumman. The results of the normal modes analysis of the FE models with different flexibility formulae were compared with those obtained from the experimental modal analysis (EMA). EMA was performed to measure the natural frequencies and mode shapes of the structure with bolted joints. The total errors calculated from the comparisons between FE and EMA were recorded for all flexibility formulae. The results of the investigation demonstrated that there are noticeable variations in the total errors recorded using different flexibility formulas for the CBUSH element representing the bolts in the FE model of the structure with bolted joints. It was found that the Swift’s flexibility formula is the most appropriate for the prediction of the natural frequencies of the structure with bolted joints. [ABSTRACT FROM AUTHOR]

The accuracy of the predicted dynamic behaviour of a structure is highly dependent on the accurate properties used in the analytical model. The predicted results computed from the finite element method used to develop the model are often found not to be in good agreement with the experimental results due to the erroneous assumptions made in the finite element modelling. One way to systematically correct the finite element model is to use model updating methods. The main goal of this study was to systematically improve the predicted dynamic behaviour of a car trunk lid using the modal based updating method. HyperWorks and Leuven Measurement System (LMS) test were used for the finite element modelling and modal testing respectively. In order to evaluate the capability of the modal based updating in improving the finite element model, the natural frequencies and mode shapes computed from the updated finite element model of the trunk lid were compared with the corresponding experimental results. The comparison of the results showed that the total error of 65.48 percent predicted from the initial finite element model of the trunk lid has been successfully reduced to 4.74 percent. In conclusion, modal based updating has been successfully used to reconcile the initial finite element model with the physical model of the trunk lid. [ABSTRACT FROM AUTHOR]

Structural dynamic modification (SDM) techniques are important in structural optimisation of the system to modify the dynamics behaviour of the structure. It can be implemented by utilising the dynamic test data to alters the dynamic behaviour of the structure. However, there are some concern in the design process in which the undesirable mode shapes movement of structure are occurred along side with desired natural frequencies. Therefore, the SDM technique can be used as to varying the predicted dynamic behaviour of structure to solve noise, vibration and harshness problems (NVH). In order to be well predicted the modified dynamic behaviour of structure, the optimisation method via finite element (FE) model updating is used to obtain a reliable and accurate unmodified FE model. The main goal of this research is to change or altered the mode shapes movement of the mild steel bar structure due to the mass modification without shifting the natural frequencies away from unmodified value. In this research, the FE model of the mild steel bar structure was constructed using MSC software and the predicted dynamic behaviour were compared with the experimental modal analysis (EMA) data. The measured data was obtained using an impact hammer testing and roving accelerometers under free-free boundary conditions with data were extracted from LMS SCADAS software. Then, reconciliation process via FE model updating method is carried out in order to reduce the inaccuracy of initial FE model based on the measured data where the initial error has been reduced from 24.81 percent to 7.71 percent. Finally, the updated FE model was used for SDM technique using mass modification. The results reveal that the mode shapes of the structure can be change by manipulating the mass. It is agreed that by adding a reasonable mass with a right position to the structure the mode shapes can be change without shifting the natural frequencies away from unmodified values. [ABSTRACT FROM AUTHOR]

A simple structure with bolted joints consists of the structural components, bolts and nuts. There are several methods to model the structures with bolted joints, however there is no reliable, efficient and economic modelling methods that can accurately predict its dynamics behaviour. Explained in this paper is an investigation that was conducted to obtain an appropriate modelling method for bolted joints. This was carried out by evaluating four different finite element (FE) models of the assembled plates and bolts namely the solid plates-bolts model, plates without bolt model, hybrid plates-bolts model and simplified plates-bolts model. FE modal analysis was conducted for all four initial FE models of the bolted joints. Results of the FE modal analysis were compared with the experimental modal analysis (EMA) results. EMA was performed to extract the natural frequencies and mode shapes of the test physical structure with bolted joints. Evaluation was made by comparing the number of nodes, number of elements, elapsed computer processing unit (CPU) time, and the total percentage of errors of each initial FE model when compared with EMA result. The evaluation showed that the simplified plates-bolts model could most accurately predict the dynamic behaviour of the structure with bolted joints. This study proved that the reliable, efficient and economic modelling of bolted joints, mainly the representation of the bolting, has played a crucial element in ensuring the accuracy of the dynamic behaviour prediction. [ABSTRACT FROM AUTHOR]

The objective of this paper is to present frequency response function (FRF) based updating as a method for matching the finite element (FE) model of a laser spot welded structure with a physical test structure. The FE model of the welded structure was developed using CQUAD4 and CWELD element connectors, and NASTRAN was used to calculate the natural frequencies, mode shapes and FRF. Minimization of the discrepancies between the finite element and experimental FRFs was carried out using the exceptional numerical capability of NASTRAN Sol 200. The experimental work was performed under free-free boundary conditions using LMS SCADAS. Avast improvement in the finite element FRF was achieved using the frequency response function (FRF) based updating with two different objective functions proposed. [ABSTRACT FROM AUTHOR]

Thin metal sheets are widely used to fabricate a various type of aerospace structures because of its flexibility and easily to form into any type shapes of structure. The riveted joint has turn out to be one of the popular joint types in jointing the aerospace structures because they can be easily be disassembled, maintained and inspected. In this paper, thin metal sheet components are assembled together via riveted joints to form a simplified model of aerospace structure. However, to model the jointed structure that are attached together via the mechanical joints such as riveted joint are very difficult due to local effects. Understandably that the dynamic characteristic of the joined structure can be significantly affected by these joints due to local effects at the mating areas of the riveted joints such as surface contact, clamping force and slips. A few types of element connectors that available in MSC NATRAN/PATRAN have investigated in order to presented as the rivet joints. Thus, the results obtained in term of natural frequencies and mode shapes are then contrasted with experimental counterpart in order to investigate the acceptance level of accuracy between element connectors that are used in modelling the rivet joints of the riveted joints structure. The reconciliation method via finiteelement model updating is used to minimise the discrepancy of the initial finite element model of the riveted joined structure as close as experimental data and their results are discussed. [ABSTRACT FROM AUTHOR]

In this paper, frequency based substructuring (FBS) is used in an attempt to predict the dynamic behaviour of an assembled structure. The assembled structure which consists of two beam substructures namely substructure A (finite element model) and substructure B (experimental model) was tested. The FE model of substructure A was constructed by using 3D elements and the Frequency Response Functions (FRFs) were derived viaa FRF synthesis method. A specially customised bolt was used to allow the attachment of sensors and excitation to be made at theinterfaces of substructure B, and the FRFs were measured by using an impact testing method. Both substructures A and B were then coupled by using the FBS method for the prediction of FRFs. The coupled FRF obtained was validated with the measured FRF counterparts. This work revealed that by implementing a specially customised bolt during the measurement of FRF at the interface, led to an improvement in the FBS predicted results. [ABSTRACT FROM AUTHOR]

In this paper, finite element (FE) joint modelling technique for prediction of dynamic properties of sheet metal jointed by tungsten inert gas (TTG) will be presented. I cross section dissimilar flat plate with different series of aluminium alloy; AA7075 and AA6061 joined by TTG are used. In order to find the most optimum set of TTG welding dissimilar plate, the finite element model with three types of joint modelling were engaged in this study; bar element (CBAR), beam element and spot weld element connector (CWELD). Experimental modal analysis (EMA) was carried out by impact hammer excitation on the dissimilar plates that welding by TTG method. Modal properties of FE model with joints were compared and validated with model testing. CWELD element was chosen to represent weld model for TTG joints due to its accurate prediction of mode shapes and contains an updating parameter for weld modelling compare to other weld modelling. Model updating was performed to improve correlation between EMA and FEA and before proceeds to updating, sensitivity analysis was done to select the most sensitive updating parameter. After perform model updating, average percentage of error of the natural frequencies for CWELD model is improved significantly. [ABSTRACT FROM AUTHOR]