TY - BOOK

T1 - Tailoring structures using stochastic variations of structural parameters

AU - van den Broek, Sander Friso

N1 - Doctoral thesis

PY - 2023

Y1 - 2023

N2 - Imperfections, meaning deviations from an idealized structure, can manifest through unintended variations in a structure’s geometry or material properties. Such imperfections affect the stiffness properties and can change the way structures behave under load. The magnitude of these effects determines how reliable and robust a structure is under loading. Minor changes in geometry and material properties can also be added intentionally, creating a more beneficial load response or making a more robust structure. Examples of this are variable stiffness composites, which have varying fiber paths, or structures with thickened patches. The work presented in this thesis aims to introduce a general approach to creating geodesic random fields in finite elements and exploiting these to improve designs. Random fields can be assigned to a material or geometric parameter. Stochastic analysis can then quantify the effects of variations on a structure for a given type of imperfection. Information extracted from the effects of imperfections can also identify areas critical to a structure’s performance. Post-processing stochastic results by computing the correlation between local changes and the structural performance result in a pattern, describing the effects of local changes. Perturbing the ideal deterministic geometry or material distribution of a structure using the pattern of local influences can increase performance. Examples demonstrate the approach by increasing the deterministic (without imperfections applied) linear buckling load, fatigue life, and post-buckling path of structures. Deterministic improvements can have a detrimental effect on the robustness of a structure. Increasing the amplitude of perturbation applied to the original design can improve the robustness of a structure’s response. Robustness analyses on a curved composite panel show that increasing the amplitude of design changes makes a structure less sensitive to variations. The example studied shows that an increase in robustness comes with a relatively small decrease in the deterministic improvement.

AB - Imperfections, meaning deviations from an idealized structure, can manifest through unintended variations in a structure’s geometry or material properties. Such imperfections affect the stiffness properties and can change the way structures behave under load. The magnitude of these effects determines how reliable and robust a structure is under loading. Minor changes in geometry and material properties can also be added intentionally, creating a more beneficial load response or making a more robust structure. Examples of this are variable stiffness composites, which have varying fiber paths, or structures with thickened patches. The work presented in this thesis aims to introduce a general approach to creating geodesic random fields in finite elements and exploiting these to improve designs. Random fields can be assigned to a material or geometric parameter. Stochastic analysis can then quantify the effects of variations on a structure for a given type of imperfection. Information extracted from the effects of imperfections can also identify areas critical to a structure’s performance. Post-processing stochastic results by computing the correlation between local changes and the structural performance result in a pattern, describing the effects of local changes. Perturbing the ideal deterministic geometry or material distribution of a structure using the pattern of local influences can increase performance. Examples demonstrate the approach by increasing the deterministic (without imperfections applied) linear buckling load, fatigue life, and post-buckling path of structures. Deterministic improvements can have a detrimental effect on the robustness of a structure. Increasing the amplitude of perturbation applied to the original design can improve the robustness of a structure’s response. Robustness analyses on a curved composite panel show that increasing the amplitude of design changes makes a structure less sensitive to variations. The example studied shows that an increase in robustness comes with a relatively small decrease in the deterministic improvement.

U2 - 10.15488/13560

DO - 10.15488/13560

M3 - Doctoral thesis

CY - Hannover

ER -