F14
Manufacturing Process Sim 1

Back to overview

17:35
conference time (CEST, Berlin)
Machining Induced Distortion Modelling of 316L Stainless Steel Using FEM
27/10/2021 17:35 conference time (CEST, Berlin)
Room: F
A. Zonuzi, T. Syed (Nuclear AMRC, University of Sheffield, GBR)
A. Zonuzi, T. Syed (Nuclear AMRC, University of Sheffield, GBR)
Part distortion is a common problem in the manufacturing life cycle and is defined as the deviation of part shape from original intent after being released from the fixture. Excessive distortion that occurs after machining creates the need for secondary corrective operations or even parts to be scrapped in extreme cases which costs money and time to the manufacturer. It was found through a literature review that a significant amount of research has been done on distortion prediction in aerospace components or materials (Aluminium and Titanium), however there was a lot less research on components and materials like stainless steel that are used in the civil nuclear industry. This paper aims to predict distortions induced during milling of 316L stainless steel through a Finite Element Model (FEM) and understand the effects of various parameters in order to optimise the process . A full factorial design of experiments was considered to review the effects of key process parameters and this was replicated both in modelling and real time experiments. A slot milling operation for a simple geometry was selected for this research. Along with cutting parameters (feed and spindle speed), clamping force of the fixture was also taken into account in simulations. The workpieces were scanned using a coordinate measuring machine (CMM) before and after machining in order to determine the distortion induced in the component. Hexagon’s MSC MARC was used to simulate the cutting process, which takes the material's heat treatment history in consideration. The software also uses the toolpath information during simulations which helps the user to achieve a more accurate model. Comparing simulated results against real time experiments, it was found that the initial residual stress balance in the material has a significant impact on distortions rather than the cutting and process parameters hence workpiece’s process history plays a critical role in distortion modelling. The main aim of this research is to develop a methodology of using FEM to predict part distortion in order to minimise or eliminate defects and successive corrective actions during machining of large high value components within the civil nuclear industry. In terms of future work, the team at Nuclear AMRC are looking to investigate various workpiece materials that are also widely used in the industry along with complex workpiece geometries and toolpaths to replicate the process on actual components.
316L stainless steel, milling, FEM, distortion modelling, residual stress
17:55
conference time (CEST, Berlin)
Integrated Process Simulation and Optimization of Injection Moulding Process
27/10/2021 17:55 conference time (CEST, Berlin)
Room: F
S. Kulkarni, D. Fust (Vanderplaats R&D DBA OmniQuest, USA)
S. Kulkarni, D. Fust (Vanderplaats R&D DBA OmniQuest, USA)
Design space exploration and optimization in injection molding is usually an ad-hoc process dependent upon the expertise of the user. Additionally, due to the lack of deterministic equations connecting the process inputs and outputs, the design space is not well understood and gradient based optimization methods cannot be directly applied. In this work, we demonstrate the improvement in the process output of an injection molding process using automated response surface modelling and optimization. This work utilizes collaborative simulation between OmniQuest™ Iliad™ Design Exploration & Automation Studio and Autodesk® Moldflow Insight®. A simulation model is first created in Moldflow Insight with process setting values based on user judgement. The input and output process parameters contained in this simulation are imported into Iliad through a dedicated interface. Next, a Latin Hypercube design of experiments is created in Iliad for five process settings (melt and mold temperatures, cooling time and gate location (X & Y coordinates)) and three outputs (maximum volumetric shrinkage, mass, and maximum clamping force) selected from the imported process settings. This results in 40 design points which are evaluated in Moldflow Insight, executed via Iliad. The resulting dataset is then used to fit a second order response surface model. An optimization study is then run using Iliad’s optimization component and the surrogate model to minimize the volumetric shrinkage of the finished part. Gradient based constrained optimization is run using the modified method of feasible direction method. This is enabled by the DOT optimization engine developed by OmniQuest. Results show a reduction of 32% in the maximum volumetric shrinkage compared to the initial solution, demonstrating improved process performance. Additionally, the process of exporting resulting plots is also automated using a Moldflow macro and the Python script component in Iliad, encapsulating the entire design workflow in a single platform.
Optimization, design exploration, simulation integration, automated design improvement
18:15
conference time (CEST, Berlin)
Efficient Cure Cycle Optimization with Recurrent Neural Network Surrogate Models
27/10/2021 18:15 conference time (CEST, Berlin)
Room: F
A. Floyd, S. Reid, A. Stewart (Convergent Manufacturing Technologies Inc, CAN)
A. Floyd, S. Reid, A. Stewart (Convergent Manufacturing Technologies Inc, CAN)
Physics-based process simulation is a cost-effective alternative to trial and error when determining suitable process parameters for the manufacture of a composite structure. A critical component of the manufacturing process is the temperature cycle imposed on the structure during cure. High fidelity 3D finite element analyses can be used to accurately predict the temperature history and subsequent material properties for a given cure cycle, but at a significant computational cost. Determining the optimal cure cycle which minimizes potential manufacturing defects and meets specifications using high fidelity finite element simulations can be prohibitively time-consuming as typical gradient-based optimization techniques may require thousands of model evaluations. In many common manufacturing scenarios, a 3D structure can be approximated by one or more 1D slices which significantly reduces the time required to run a simulation, but it still may take hours to complete the optimization. This work demonstrates how recurrent neural network (RNN) surrogate models can be used as a computationally inexpensive alternative to finite element analyses for cure cycle optimization. RNN surrogate models have been trained for predicting process parameters such as temperature, degree of cure, and resin viscosity of 1D “drill-through” stacks of larger assemblies consisting of a composite laminate on a tool. Leveraging the speed of these surrogate models, they have been subsequently used in an optimization algorithm to determine the shortest duration cure cycle that meets the manufacturing requirements. Multiple 1D stacks can be considered in parallel to ensure the cure cycle is valid for the whole assembly and not just a single location. The manufacturing requirements are defined by the process specification, which typically changes from process to process and this method is shown to be capable of handling process specifications of “real-world” complexity. Ultimately, it is shown that the time required to perform the optimization using the surrogate model is considerably less than the time required to run the optimization using the finite element model and both methods converge to comparable optimized cure cycles.
composites, process simulation, optimization, neural networks, surrogate model, machine learning
18:35
conference time (CEST, Berlin)
Linking Manufacturing Process to Advanced Material Performance Using Multiscale Modeling
27/10/2021 18:35 conference time (CEST, Berlin)
Room: F
H. Cornwell (Siemens Digital Industries Software, USA)
H. Cornwell (Siemens Digital Industries Software, USA)
The introduction of advanced materials to established and emerging industries has allowed companies to optimize product performance and maintain a competitive edge. Although these new materials may have favorable mechanical and thermal properties compared to traditional materials, the manufacturing processes are often complex and very sensitive to processing parameters. In fact, material properties are directly related to the manufacturing process and need to be accounted for when analyzing a part’s performance. This paper will outline four different advanced materials and manufacturing processes that demonstrate how combining process simulation with microstructural multiscale modeling allows for engineers to accurately characterize material properties. The first case study investigates injection molded parts with short fiber filled polymers and the effect the complex fiber orientation distribution have on the mechanical response. Next, the mechanical properties of a laminate textile composite component are compared against simplistic materials models when accounting for the tow shearing caused by draping or forming. The next study evaluates the high volume manufacturing process of sheet molded composites by evaluating strength variation due to randomly aligned chopped unidirectional prepreg tapes. Finally, repeating unit cell structures, possible through additive manufacturing, allow for customized elastic and failure properties in light weighted parts. Each of these case studies demonstrate how manufacturing processes information is utilized to create microstructural finite element models (FEM) that can be mapped to a component model for a fully coupled multiscale analysis for an accurate prediction of the part’s performance. While manufacturing induced material variation is traditionally seen as a phenomenon that needs to be mitigated, these studies demonstrate how material variation can be used to improve part performance. Multiscale modeling proves to be a flexible and robust method for connecting the manufacturing process to part analysis in an automated or streamlined workflow and is critical for bringing optimized and reliable products to market.
multiscale, materials, composites, finite element analysis, manufacturing variation
×

[TITLE]

[LISTING

[ABSTRACT]

[DATE]

[ROOM]

[KEYWORDS]