License Authors of papers retain copyright and release the work under a Creative Commons Attribution 4.0 International License (CC-BY).

An automated reverse engineering process is developed that uses a structured light optical measurement system to collect dense point cloud geometry representations. The modeling process is automated through integration of software for point cloud processing, reverse engineering, solid model creation, grid generation, and structural solution. Process uncertainties are quantified on a calibration block and demonstrated on an academic transonic integrally bladed rotor. These uncertainties are propagated through physics-based models to assess impacts on predicted modal and mistuned forced response. Process details are discussed and recommendations made on reducing uncertainty. Reverse engineered parts averaged a deviation of 0.0002 in. (5 ¡xm) which did not significantly impact low and midrange frequency responses. High frequency modes were found to be sensitive to these uncertainties demonstrating the need for future refinement of reverse engineering processes.

The impact of geometry variations on integrally bladed disk eigenvalues is investigated. A large population of industrial Bladed Disks (Blisks) are scanned via a structured light optical scanner to provide as-measured geometries in the form of point-cloud data. The point cloud data is transformed using Principal Component Analysis that results in a Pareto of Principal Components (PCs). The PCs are used as inputs to predict the variation in a Blisk’s eigenvalues due to geometry variations from nominal when all blades have the same deviations. A large subset of the PCs are retained to represent the geometry variation, which proves challenging in probabilistic analyses because of the curse of dimensionality. To overcome this, the dimensionality of the problem is reduced by computing an active subspace that describes critical directions in the PC input space. Active variables in this subspace are then fit with a surrogate model of a Blisk’s eigenvalues. This surrogate can be sampled efficiently with the large subset of PCs retained in the active subspace formulation to yield a predicted distribution in eigenvalues. The ability of building an active subspace mapping PC coefficients to eigenvalues is demonstrated. Results indicate that exploitation of the active subspace is capable of capturing eigenvalue variation.

The impact of geometry variations on integrally bladed disk eigenvalues is investigated. A large population of industrial bladed disks (blisks) are scanned via a structured light optical scanner to provide as-measured geometries in the form of point-cloud data. The point cloud data are transformed using principal component (PC) analysis that results in a Pareto of PCs. The PCs are used as inputs to predict the variation in a blisk's eigenvalues due to geometry variations from nominal when all blades have the same deviations. A large subset of the PCs is retained to represent the geometry variation, which proves challenging in probabilistic analyses because of the curse of dimensionality. To overcome this, the dimensionality of the problem is reduced by computing an active subspace that describes critical directions in the PC input space. Active variables in this subspace are then fit with a surrogate model of a blisk's eigenvalues. This surrogate can be sampled efficiently with the large subset of PCs retained in the active subspace formulation to yield a predicted distribution in eigenvalues. The ability of building an active subspace mapping PC coefficient to eigenvalues is demonstrated. Results indicate that exploitation of the active subspace is capable of capturing eigenvalue variation.

No abstract