Increasing computing powers, advanced software, and smart technical devices for collecting and processing manufacturing data have turned the Digital Twin (DT) concept into a practical instrument for real-time geometry assurance. However, implementing the DT for geometry assurance in industry requires selecting an appropriate infrastructure that guarantees a seamless flow of geometry and deviation information across design, production, inspection, and variation simulation. The Quality Information Framework (QIF), an inspection-focused standard for Model-Based Definition, can address this challenge and close currently open gaps in the Digital Thread for geometry assurance. This article proposes an operational integration framework that uses QIF to set up the DT for non-rigid variation simulation. An inverse mapping algorithm divides Finite Element meshes into subsets representing the QIF features in variation simulation. The identified features connect the simulation model with the measurement information, which is directly embedded in QIF and linked to the features. Measured size, location, orientation, and form deviations of a batch of parts are exchanged collectively via QIF, automatically mapped on shell meshes, and used for DT applications without losing the link to the physical counterparts. An exemplary implementation of the framework and its application to find optimal part matching for a sheet metal assembly emphasizes the potential of QIF for feeding the DT.

In high-volume automotive production there is a need to describe the geometry of a part more accurately in early product-realization phases, when comprehensive measurement data are not yet available at scale, so that large-scale statistical variation analyses can be run quickly with simplified yet credible models. This paper proposes a method to predict part shape from sparse information by optimizing both the number and placement of inspection points subject to a defined accuracy threshold. Two production body-in-white parts, a compact inner d-pillar and a large outer fender, were 3D scanned with a Nikon H120 blue-laser high-precision scanner and compared against finite-element (FE) simulation models. The surface is reconstructed by combining the mesh and stiffness matrix of an FE-based material model implemented in a computer-aided tolerancing (CAT) software, RD&T, and the method is commissioned/validated against the scan. The study is experimental: a high-density scan is used as reference to evaluate how accurately sparse point constraints can reconstruct the full-field deformation. An error-driven greedy algorithm is used to iteratively select additional inspection points at locations where the current reconstruction deviates most from the scanned geometry, until the accuracy criterion is satisfied while minimizing model size and computation time.Results indicate a correlation between the 3D-scanned geometry and the simulation-based reconstructions, demonstrating that the approach can reproduce part shape with reasonable precision for early-phase analyses. The method enables rapid simulations on large populations of parts with a reduced model that remains accurate enough to support decision-making in tolerancing and variation analysis. The reduced model is compatible with standard non-rigid Monte Carlo variation simulations in RD&T, enabling fast early-phase tolerance studies on large virtual populations of parts.

The Quality Information Framework (QIF) standard aims to cover the whole product quality ecosystem using a Model-Based Definition approach. Compared to the STEP AP242 and JT standard, QIF concentrates on semantically intertwining information from product design and inspection within one information carrier. Due to its strong focus on geometrical product quality, QIF is predestined to be used as the basis for activities that aim to assure product quality in the presence of geometrical variations. Still, its potential for variation simulation, virtually predicting the effects of geometrical part and assembly variations on the final product quality, remains largely unexploited. Methods for bringing information from QIF to variation simulation and translating it into deviations are currently lacking. This article presents a novel QIF-based method that fosters the automation of deviation modeling in non-rigid variation simulation of sheet metal assemblies. Building upon an inverse mapping method for registering QIF features within shell meshes leads to a hybrid model that combines the strengths of a parametric and discrete geometry representation. By exploiting this duality, size, location, orientation, and form tolerances are translated into mesh deviations using a joint Small Displacement Torsor and Skin Model Shapes approach. An exemplary application of the proposed method to a spot-welded sheet metal assembly illustrates QIF’s potential for bridging feature-based GD&T specification and mesh-based variation simulation.

Most metallic components undergo machining to some extent during manufacturing. It is a very complex process where extreme temperatures and forces are generated on the surface, primarily due to friction. These conditions often give rise to significant residual stress on the surface of the finished product, which may be sufficient to alter both the behavior of the material and the geometry of the product. The geometric effects of these stresses are further complicated by the presence of residual stresses from previous processes, such as casting or additive manufacturing (AM). A methodology for applying machining-induced residual stress onto a geometry to estimate the geometric distortion that can be expected from a machining process in an AM context is proposed. This study compares a range of feasible residual stress magnitudes to demonstrate the importance of tailoring the process parameters to minimize the effects of machining induced residual stress on the geometric variation.

ABSTRACT: Colour, Material and Finish (CMF) designers face rising circularity demands but lack tools that combine reliable data, traceable reasoning and creative control. This paper reports a case study with automotive CMF designers, identifying pain points in data access, evaluation of circular options, authorship and trust in AI. We propose design requirements and a conceptual model for agentic AI systems that support circular CMF work while preserving designer agency, accountability, and confidence in material decisions.

Variation simulation is widely used for statistical quality control of mechanical assemblies subject to geometric variation. A major challenge lies in accurately modeling contact interactions between rigid components to consistently configure the rigid bodies. This paper presents a Monte-Carlo variation simulation framework that explicitly incorporates rigid-body normal contact using unilateral non-penetration constraints. The proposed method is applied to two reference assemblies and compared against a conventional contactless variation simulation approach. Results show that enforcing contact constraints significantly alters the predicted assembly behavior by reducing the 6σ spread of critical measurement points, while introducing a directional bias along the contact normal reflected by a shift in the mean displacement. These findings demonstrate that contact modeling plays a critical role in realistic variation simulation. The proposed framework provides a physically consistent and computationally efficient basis for contact-aware tolerance analysis of rigid-body assemblies.

Computer-Aided Tolerancing (CAT) software has become the standard for statistically analyzing the effects of geometrical part variations on product quality. Irrespective of CAT's scope and technical depth, Finite Element Analysis (FEA) software, used to simulate the physical product behavior for ideal part geometry in the first place, is also often used for studies with geometrical shapes deviating from their nominal. However, this requires a manual translation of the tolerances specified in the design phase into geometrical variations represented by Finite Element (FE) meshes and their transfer to the FEA software. The method presented in this article exploits the potential of Model-Based Definition by establishing a link between Computer-Aided Design and FEA software to empower the latter for variation simulation based on semantic Geometric Dimensioning and Tolerancing (GD&T) information. To transfer this information exchanged via the Quality Information Framework (QIF) standard, a new mapping algorithm is presented that automatically decomposes FE meshes into geometrical face elements and creates a semantic link with the GD&T information carried in QIF. As a result, geometrical features are simultaneously described through meshes with nodes in the 3D Euclidean space and mathematical geometrical faces in the 2D parameter space. Exploiting this duality, mesh deviations are modeled indirectly by adjusting the mapped feature descriptions. An exemplary implementation in ANSYS (R) and its usage for non-intrusive structural simulations illustrates that sharing tolerancing information via QIF enables an automated, GD&T standards-compliant variation simulation within FEA software environments and is one step closer to a seamless digital thread for geometry assurance.

In the assembly of large carbon fiber reinforced polymer (CFRP) structures, geometric deviations create interface gaps that must be filled with shims before fastening to avoid forced-fit stresses. The design gap, the interface clearance intentionally built into the nominal geometry, governs both the shimming demand and the post-assembly geometric deviation, yet it is currently specified by engineering judgment rather than analysis. If the design gap is too small, contact forces deform the parts and increase the deviation after fixture release. If it is too large, the shim thickness grows, adding weight and fastener bending stress. This paper treats the design gap as an explicit design variable in a non-rigid variation simulation and sweeps it across two populations of 100 virtual CFRP wingbox assemblies with different manufacturing signatures, using the method of influence coefficients with contact modeling. Geometric deviation decreases asymptotically with increasing design gap while shim thickness increases monotonically. For the wingbox studied, the deviation RMS falls from a baseline of 0.349 mm to a minimum of 0.247 mm at a design gap of approximately 0.3 mm, beyond which shim thickness keeps growing with no geometric benefit. The improvement depends strongly on the manufacturing signature, ranging from 29

In the assembly of large-scale composite structures, interface gaps arising from part deviations are commonly filled with shims to avoid induced stresses due to the forced contact. However, excessive shim thickness increases bending stresses in fasteners, while the process of gap minimization itself can distort the final assembly shape. This study presents a framework that integrates fastener load limits with fixture optimization to determine maximum allowable shim thickness and reduce interface gaps. Using a combination of finite element simulation, contact modeling, and numerical optimization, the method computes optimal fixture adjustments to minimize gaps while respecting structural constraints. A case study on a simplified wingbox section shows that, although local gaps are reduced by an average of 43%, global geometric deviations often increase, revealing a fundamental trade-of between gap minimization and geometric accuracy. The framework supports pre-manufactured shims and is particularly suited for digital-twin environments, enabling inspection data to inform fixture setup before assembly. Results highlight the importance of jointly considering structural and geometric objectives and suggest that multi-objective optimization and alternative shimming strategies can further enhance geometry assurance and joint integrity.

Ensuring geometric accuracy in joining mechanical structures is a key challenge in smart manufacturing, particularly during assembly. To improve assembly and joining accuracy, this paper employs digital twin technology, with a focus on high-fidelity digital modeling as a foundational step toward realizing digital twins–while treating bidirectional automated data flow as a future development direction. Digital twin models are developed for two representative joint types: threaded connections and spot-welded sheet metal structures, incorporating critical influences from manufacturing, assembly, and service phases. Compared to traditional idealized simulations, the proposed models enable more effective assembly process optimization and enhanced geometric quality control.

Megacasting enables the production of large, structurally integrated aluminum components for automotive applications, yet the geometrical distortion remains a major challenge. This study combines industrial observations, experimental trials, and literature analysis to identify and interpret key process parameters affecting distortion in high-pressure die casting of large components. The work integrates quantitative measurements from an industrial case study with qualitative input from experienced foundry engineers. The results highlight how die temperature distribution, cooling channel placement, solidification and cooling times, and quenching strategies influence residual stresses and distortion. Stress relaxation at elevated temperatures and constraint removal during ejection are highlighted as key mechanisms contributing to distortion formation.Building on these insights, potential approaches such as tool geometry correction and simulation-based prediction of distortion are discussed with respect to industrial feasibility. The findings contribute to improved understanding of distortion phenomena in megacasting and support future research toward predictive modeling and compensation strategies.

In order to reduce fuel consumption, aero engines are increasingly made from complex welded assemblies rather than singular monolithic castings. This increases variability in the manufacturing process, calling for an advanced geometry assurance approach to ensure adequate quality. Digital twins for geometry assurance have previously been suggested as way of increasing the precision of welded assemblies. However, these systems have mostly remained at the conceptual or experimental level, and their potential has not been realized in the industrial environment. No clear and unanimous view exists for industrial implementation of digital twins for manufacturing. This paper proposes a digital twin for welded assemblies, combined with a roadmap for industrial implementation based on established industry standards. The proposed digital twin functionality is based on individual locator adjustment combined with a genetic optimization algorithm and a non-nominal weld simulation method. This functionality is organized into the ISO 23247 framework to facilitate the information flow within the digital twin and support efficient implementation. To ensure compatibility with the industrial environment, an extended activity model based on the widely used ISA-95 standard is introduced based on the activities defined within the ISO 23247 framework implementation. This extended activity model for digital twins has a networked structure, enabling fast and direct information exchange between different activities while maintaining compliance with the ISA-95 information model. This lays the foundation for a digital thread to support the digital twin with relevant data. By combining the digital twin with existing frameworks, a full methodology is formed for the industrial implementation and realization of the proposed digital twin for welded assemblies. Finally, a case study shows how a digital twin implementation in Matlab and RD T can decrease variation in a welded assembly. This methodology is applicable for implementation of a wide range of digital twin functionalities within industrial environments based on the ISA-95 standard.

Residual stress plays an important role in modern manufacturing, especially with the increasing change of manufacturing environment with mega-casting and additive manufacturing. Advanced manufacturing processes introduce several constraints, and one critical constraint to be predicted and controlled is residual stress, which could affect the geometrical quality during assembly. The stresses might lead to displacement and reduce the deviations of the final product. To ensure this, residual stresses are to be measured, predicted, and controlled throughout part manufacturing and assembly. In this paper, the impact of residual stress on geometrical deviation and variations is investigated through a case study: a small-scale cast bracket structure. A combined approach of casting simulation and variation simulation is applied to quantify the residual stress and to integrate, and identify the influence of residual stress during assembly in geometrical variations. The simulation results emphasize the role of residual stress in the deformations and the tolerances of the assembly and cast parts. This research demonstrates and also contributes methods for the integration of residual stress and geometry assurance to ensure more reliable assemblies in advanced manufacturing.

Friction at contact interface strongly influences deformable assembly variation, yet most nonrigid variation simulation tolerancing workflows still assume frictionless contact. This paper presents a novel nonlinear variation simulation and unilateral contact considering Coulomb friction solved as a second order cone program, enabling nonlinear tolerancing with practical computational cost. The method is demonstrated on an industrial assembly with perturbed geometries. Compared to a frictionless baseline, the proposed model captures stick and slip effects and yields lower tangential slip and reduced variation across defined contact zones. The results highlight the necessity of frictional contact modeling for tolerance analysis in assemblies.

Optimization-based digital human modelling (DHM) can compute manikin motions for unique work tasks, requiring no motion capturing or motion data manipulation to simulate new work tasks. Also, optimization-based DHM can consider prevailing force and torque exertions, e.g. pushing or twisting, in the motion computations. This makes optimization-based DHM well suited for assessing workstation designs early in virtual development phases. When using optimization-based DHM to simulate work tasks and determine task times in settings such as manual assembly, it is crucial that the manikin motion durations can be set to comply with predetermined motion time systems (PMTS) data. As part of realizing this objective, this study compares assembly times generated by an optimization-based DHM tool, where durations of discrete manikin motions are determined based on PMTS data, against an industrial use case with known assembly times, determined according to the company standard. The comparison aims to identify the differences between the times generated by the DHM tool and the times determined in accordance with the company standard, understand why they occur and how they potentially can be addressed. The findings support establishing a road map for future research and development for improving task time estimations in optimization-based DHM.

This paper presents a novel digital twin framework employing batch incremental learning for geometry assurance. Addressing quality issues caused by part and process variation, the method evaluates three critical tasks: part matching, locator adjustments, and joining sequence. The proposed framework utilizes deep learning architectures, each trained on recursive simulation data. Employing incremental learning, the models adapt to new batch characteristics while maintaining predictive accuracy. A spot welded assembly demonstrated the proposed approach efficiency, achieving prediction accuracies with errors as low as 0.02 mm for part matching and 0.1 mm for locator adjustments. (c) 2025 The Author(s). Published by Elsevier Ltd on behalf of CIRP. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)

In the machining industry, fixturing of indexable inserts in the tool holders plays a critical role in ensuring high-quality outcomes by minimizing insert movement during operations. Ensuring consistent pressure distribution in an indexable cutting tool interface is essential for extending the lifespan of a carbide insert. However, existing methods either lack the necessary complexity to accommodate varying loads or are overly intricate for implementation in the early stages of product development. To address this gap, a novel approach was developed that integrates the contact index algorithm into a robust locating scheme optimization. The results show that it is possible to design an indexable cutting tool where clearly defined contacting points (half-spheres) can support and maintain minimal movement in an indexable insert during its expected lifetime.

Robust Design is essential for developing high-quality products by minimizing their sensitivity to variation and uncertainties from diverse sources. Despite the wide range of approaches from industry and research, it faces challenges due to their siloed usage. Motivated to gain a better understanding of these challenges and to derive implications for how to overcome them, this paper discusses different viewpoints on Robust Design, taking into account different life cycle phases, domains, and system levels. This highlights the challenges of enhanced product complexity and shifting focuses, for instance, when it comes to cyber-physical systems or sustainable design, and the need for collaboration. Accordingly, the vision of a collaborative approach to Robust Design in a team is presented, in which the actors’ strengths are combined to further increase efficiency and significance.

An essential premise fora reliable variation simulation is that information on the geometrical part variations and their accumulation and propagation within an assembly is available, accessible, interchangeable, and usable in all geometry-related downstream activities. For this reason, this article studies the potential of the QIF (Quality Information Framework) standard. It illustrates how it can be used in the sense of Model- Based Definition to close gaps in the digital geometry assurance process. Besides benefits in the automation of variation simulation, it demonstrates that the semantic, feature-based linkage between product specification and inspection information in QIF 3.0 facilitates the augmentation of variation simulation with more detailed feature information for pre-production applications and feeding the digital twin to assure and optimize product quality in the production phase.

Variation simulation approaches are frequently used to analyse the effects of geometrical variations on the final product quality. Various software tools are used during product development as they strongly differ in their specified goals, the context of use, and users. Although a few workarounds and information-sharing strategies exist, switching software usually results in the simulation model being built from scratch, leading to redundant manual effort and uncertainties. This paper examines the potential and limitations of the Quality Information Framework (QIF) information model in improving collaborative work within a heterogeneous simulation software landscape by exchanging variation simulation model-related information in a standardised Model-Based Definition sense. An application scenario shows how QIF can bridge the gap between tools used in early and late design phases.