A Unified Framework for Integrated Thermal-Fluid-Structure Modeling in PDM Systems

Thesis Abstract

This study addresses the fragmentation of traditional thermal, fluid, and structural analyses in product data management (PDM) environments by proposing a unified framework that integrates multi-physics modeling within enterprise data workflows to enable concurrent design optimization. The problem arises from siloed simulation tools, inconsistent data representations, and limited interoperability across engineering domains, which hinder rapid decision-making and accurate prediction of coupled phenomena in complex systems. The aim is to develop and validate a holistic modeling framework that (i) harmonizes data schemas and ontologies across thermal, fluid, and structural domains, (ii) embeds multi-physics coupling within a PDM-based architecture, and (iii) facilitates decision-support through integrated analytics and visualization. Specific objectives include (1) to formalize a domain-agnostic data model and interoperability protocol for thermal-fluid-structure simulations in PDM; (2) to implement a modular multi-physics coupling mechanism with dynamic load path tracking; (3) to develop calibration and validation procedures using a representative set of engineering components; (4) to evaluate the framework through a case study on a heat exchanger assembly under transient thermal loading; and (5) to quantify improvements in design cycle time and predictive accuracy relative to conventional disjoint workflows. The methodology adopts a mixed-methods design anchored in a theoretical blend of multiphysics coupling theory and information integration theory. The population comprises engineering design teams and computational analysts in a manufacturing research facility, with a purposive sample of 24 engineers and 8 researchers across mechanical, thermal-fluid, and materials specialties. Data collection instruments include structured interviews and time-motion logs to capture workflow patterns (n=32 sessions), a bespoke multi-physics simulation suite to collect computational outputs (yielding approximately 1,200 high-fidelity simulation runs across scenarios), and a standardized survey to assess usability and perceived decision-support value (n=120 respondents). Validity and reliability are addressed via triangulation of interview data with empirical workflow metrics and cross-validation of simulation results against experimental benchmarks. The data analysis plan comprises (i) regression analysis and ANOVA to evaluate the impact of the unified framework on design cycle time and prediction accuracy; (ii) Bayesian calibration to update model parameters with observational data; (iii) sensitivity and uncertainty quantification using Sobol indices for coupled thermal-fluid-structural responses; (iv) graph-based analysis to elucidate data dependencies and workflow bottlenecks; and (v) thematic analysis of interview transcripts to identify barriers to adoption and organizational fit. The model specification includes a hierarchical data-entity-relationship schema and a coupling engine implementing impedance-mmatched exchanges for temperature, pressure, stress, and displacement fields, with explicit traceability to PDM objects and revision histories. Key expected findings include demonstrable reductions in design iteration time by 28–35% and improvements in predictive accuracy of coupled responses by 15–22% compared with baseline decentralized workflows; evidence that the integrated coupling engine preserves data provenance and enables reproducibility across teams; and identification of critical data-quality requirements and alignment needs for successful cross-domain collaboration. The study anticipates that the framework will enable practitioners to perform rapid what-if analyses and concurrently optimize thermal, fluid, and structural performance within a single PDM-enabled environment, supported by a formalized mapping between physics-based models and enterprise data schemas. Contributions to knowledge include (i) a formalized, domain-agnostic data model and interoperability protocol for integrated multi-physics simulation in PDM contexts; (ii) a modular coupling architecture with traceable data lineage and reproducibility guarantees; (iii) an empirical benchmark and validation strategy for coupled thermo-fluid-structural problems in product design; and (iv) empirical evidence on organizational and process factors influencing adoption of integrated modeling in engineering practice. The main conclusion anticipates that a unified framework embedded in PDM systems enhances concurrent engineering capabilities and design reliability, with recommendations emphasizing standardization of data semantics, governance of model validation, and targeted training to promote cross-domain collaboration. Practical recommendations include developing industry-wide ontologies for thermo-fluid-structural data, adopting open standards for simulation data exchange, and integrating automated verification pipelines to sustain quality assurance across product lifecycles.