No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

The ultimate strength assessment of steel tubular members is of the utmost importance to the design and maintenance of many structures including large offshore platforms. Ultimate strength assessments I using numerical solutions must model both nonlinear material and geometric behaviour. The latter must consider large displacements, very often large rotations, and possibly even large strains. These numerical solutions must be computationally efficient and be capable of running on generally available computer hardware, i.e. minicomputers. To achieve this efficiency, attention must be paid to programming considerations, and a new suite of data management modules has been developed and is described in this thesis, which minimise disk storage and speed program development. In addition, the structural modelling was carried out almost exclusively using the Semiiaaf thin shell element. The work described in this thesis considers most of the components which contribute to the numerical ultimate strength analysis of steel tubular members. Theoretically, attention has been focused in two areas, namely the geometric nonlinearity and the automatic solution of the resulting nonlinear equations. A detailed study has been carried out to understand fully the main methods of accounting for geometric nonlinearity from fundamentals of continuum mechanics. The study has considered both the Green-Lagrange and Logarithmic strain measures with a Total Lagrangian, Updated Lagrangian and Eulerian description of motion. These formulations have been included in the Semiiaaf shell element, firstly using a continuum mechanics based approach, and secondly using the more orthodox stress resultant approach. At all stages within the thesis attention is drawn to the effects of the approximations which have been made and their resulting limitations in the respective formulations. The solution of the nonlinear equations is also covered in detail using Newton-type algorithms coupled with line searches. The solution algorithms have been derived for a constrained environment where a modified version of the generalised arc-length constraint has been used. The inclusion of material nonlinearity has been well developed previously but has been included for completeness. To demonstrate the performance and limitations of the theory presented, several carefully chosen numerical examples have been included which include the analysis of tubular steel T and X joints connections and residual strength assessment of a dented pipeline riser. Where possible, results have been compared with experimental tests. The thesis concludes that for general engineering structures, the Total Lagrangian approach based on the stress resultant model gives good engineering results, even in the presence of moderately large rotations. Of the alternative formulations the Updated Lagrangian layered approach is probably the most effective for large rotations and small to moderate strains.

Digitized by Kansas Correctional Industries

Doctor of Philosophy / The aims of the Thesis were: to investigate and quantify the factors influencing the production of maximum isometric grip strength force in a sample of Australian teenagers when using JamarTM-like handgrip dynamometers; to determine the reliability of this measure over long and short retest intervals; to establish a database of anthropometric and strength values for this group and prediction equations for premorbid strengths to aid assessment of recovery in those with upper limb pathologies. The history of these handgrip dynamometers demonstrates that they have been employed in one form or another for over three hundred years and are still widely used today in hand rehabilitation and medical examinations. Many new types of dynamometers have been constructed subsequent to the ubiquitous JamarTM and have all been briefly reviewed here. Handedness (dominance) was thought to be a possible factor influencing grip strength performance and was later evaluated. But first, the Edinburgh Handedness Inventory was tested with 658 teenagers and 64 adults and confirmed to be a valid tool for assessing handedness. Its validity was improved by substituting the tasks of sweeping and opening the lid of a box for hammering and use of a screwdriver. Its excellent reliability (ICC = .78, p < 0.01) was confirmed with 45 teenagers and 45 adults. There were 235 teenagers who performed maximal isometric grip strength tests and from the results a local database was created. It was confirmed that the grip strength difference in males and females becomes significant after the age of 13 years, and that the average teenaged male is stronger than the average teenaged female by 11.2 Kg force (p < .01). Height, weight, BMI, hand dimensions, past upper limb injuries, degrees of handedness and exercise levels were measured and compared with known norms to establish that the grip strength tested sample of teenagers was representative of urban teenagers in Australia. The influence of handedness on maximal grip strength in dominant and non-dominant hands was unable to be completely ascertained due to the vast majority of the sample of teenagers being right-handed. Only 13 of the 235 teenagers used their left hand for most tasks, with another 20 using their left hands for a small majority of tasks. There was a grip strength bias towards the dominant hand of 2.63 kg force (p < .01). The most accurate way to predict the grip strength of one hand is by knowing the grip strength of the other hand. Prediction models found that 90% (R2 adj .902) and 70% (R2adj .702) of the variance in one hand could be accounted for by the grip strength of their other hand for male and female teenagers, respectively. Prediction equations were also created to assist in estimating the pre-morbid grip strength of teenagers suffering from bilateral hand injuries. If for the males, measurements for height and hand surface area were entered into these models, the grip strength of the dominant and non-dominant hands could be estimated with 62.6 and 63.5% of the variance between the real and predicted scores accounted for, respectively. For the females the prediction models using height and hand surface area could only account for 33.9 and 42.8% of the variances, with no other independent variables improving the prediction equations. The reliability of the maximal grip strength performance of 154 of these teenagers was retested after one or four weeks. A number of sub-group permutations were created for age, gender, retest time interval and handedness groups. The measures of grip strength for males were highly reliable with ICC (3,1) values ranging from .91 to .97. These measures were significantly higher than that obtained from the females, where reliability values ranged from .69 to .83. Handedness played a significant part in grip strength reliability. The dominant hand of right-handed teenagers achieved an ICC (3,1) of .97, as contrasted with the non-dominant hand of left-handers who attained a very poor ICC (3,1) of .27. The shape of the hands of the males did not influence their grip strength or their reliability values, which ranged from .954 to .973. The shape of female hands did not affect their ability to generate maximal grip strength, only its reliability. The females with hands shaped squarer-than-average had mean grip strength reliability values of ICC (3,1) at only .48, in contrast to those with longer-than-average hands who achieved a mean ICC (3,1) of .92. The handle shape of the dynamometer may disadvantage square-handed females, and this should be further investigated.