Simulation-Driven Design, Engineering & Architecture
Presentations at CDFAM Past and Future
Simulation-driven design has matured into a foundational pillar of advanced engineering and computational workflows, spanning applications from 3D printed architected materials to full-scale architectural systems.
The presentations collected here reflect how simulation is now used not only for validation, but also as a generative engine—informing, constraining, and automating decisions in part design, process planning, and system-level performance.
From real-time cloud-based multiphysics and field-driven design in industrial AM, to uncertainty-aware structural optimization and urban energy analysis, this collection showcases the depth and breadth of simulation’s role across disciplines, scales, and sectors.
“Real-time simulation is finally giving us feedback during design, not just after.”
Andreas Vlahinos
CDFAM Computational Design events create a critical environment and feedback loop between academia, software developers, and industry engineers, ensuring alignment between research potential and practical application.
Without this communication, teams risk ‘bringing a dead bird to the door’—building tools or methods no one can or will use. At CDFAM, researchers, toolmakers, and practitioners converge to build a shared foundation for simulation-driven design that is both visionary and viable.
What follows is a cross-section of presentations from CDFAM events—spanning academic research, commercial software development, and industrial application.
Each reflects a different angle on how simulation is transforming design and engineering processes, and how the conversation across these domains ensures relevance, functionality, and adoption.
“We are working toward geometries that are automatically de-homogenized and manufacturable, without the need for visual tweaking or interpretation.”
Ole Sigmund, DTU
Simulation Research and Academic Contributions to Engineering Design at CDFAM
- Sandia National Laboratories – Jeremy Lechman
Leveraging physics-based modeling for AM process optimization
Lechman presented how multiscale particle and continuum simulations are used to evaluate microstructure evolution, porosity, and thermal transport in additively manufactured components.
DTU – Ole Sigmund
Perfect Designs for Imperfect AM: topology optimization & de-homogenization
Sigmund demonstrated how de-homogenization techniques convert periodic topology-optimized microstructures into continuous solid geometries that account for process variability in AM.
- SUPSI – Alberto Ortona
Additive Manufacturing of Ceramics Using Computational Design
Ortona demonstrated how computational workflows and simulation driven design were applied to engineer support-free, thermomechanically stable ceramic parts using extrusion-based AM, with focus on toolpath-driven geometry conditioning and printable segmentation strategies tailored to the brittle behavior of ceramics.
- University of Edinburgh – Gabrielis Cerniauskas
Machine learning for lattice and metamaterial optimization
Cerniauskas presented a data-driven approach to generate high-performance lattice morphologies using supervised ML models trained on FEA simulations.
- NYU – Daniele Panozzo
ROBUST GEOMETRY PROCESSING FOR PHYSICAL SIMULATION AND SHAPE OPTIMIZATION
Panozzo introduced an end-to-end geometry processing pipeline—featuring an open-source tetrahedral mesher—designed to robustly convert broken, non-manifold, or degenerate CAD into analysis-ready domains for use in nonlinear elasticity, contact modeling, and differentiable PDE-constrained optimization workflows.
Simulation Software and Computational Design Tools
- nTop + Siemens Energy – Bradley Rothenberg & Markus Lempke
LEVERAGING COMPUTATIONAL DESIGN WITH NTOP TO DRIVE THE ENERGY TRANSITION WITH SIEMENS ENERGY
Demonstrating workflows where scalar and vector field data—such as temperature, stress, and stiffness maps—are used to drive implicit geometry creation for heat exchanger and turbine component design, enabling high-performance, multi-zone structures with field-specific material distributions and manufacturability constraints embedded directly into the design process.
- Intact Solutions
SIMULATION-DRIVEN CONTINUOUS ENGINEERING: INTACT SOLUTIONS
Presented a continuous analysis framework where meshless solvers provide design-time simulation feedback within a unified geometry kernel, enabling rapid iteration and embedded validation for applications such as customized mechanical parts, structural components, and evolving product assemblies.
- ToffeeX – Thomas Rees
INDUSTRIALIZING PHYSICS-DRIVEN GENERATIVE DESIGN -TOFFEEX
Rees detailed the integration of adjoint-based multiphysics solvers for early-stage generative exploration in fluid, thermal, and structural domains, with applications ranging from heat exchangers and microfluidic devices to pressure vessels and complex cooling geometries, where real-time feedback allows engineers to balance competing performance objectives.
- Additive Flow – Alexander Pluke
Volumetric parameter control and simulation-aware optimization
This presentation demonstrated how voxel-based simulation outputs, including stress, heat flux, and density fields, are mapped directly to manufacturing parameters like beam power and scan speed. Using the 3MF volumetric extension, this performance data is communicated seamlessly to slicing and build preparation software, enabling localized control for applications such as heat exchangers, orthopedic implants, and load-bearing brackets with optimized internal architectures.
- Altair – Joseph Flynn
THE CHALLENGES OF USING AI IN COMPUTATIONAL ENGINEERING
Altair presented tools enabling designer control over topology, lattice density, and material distribution through implicit solid modeling workflows, incorporating real-time field evaluations to predict structural performance and inform print feasibility, enabling use cases such as lightweight orthopedic devices, cellular implants, and manufacturable lattice components tuned for strength and flexibility.
- SimScale – David Heiny
Enhancing Computational Design with Real-Time Design Insights Using AI & Physics Simulations in the Cloud
SimScale’s GPU-accelerated solvers and AI surrogate models are used to run real-time CFD and thermal simulations in a web browser, enabling engineers to iteratively test and optimize heat sink and enclosure designs without local hardware, as demonstrated in an electronics cooling use case.
- Coreform – Matthew Sederberg
COREFROM: ISOGEOMETRIC ANALYSIS TO RUN FEA SIMULATIONS DIRECTLY ON FULLY-FEATURED GEOMETRY
Coreform explained how smooth T-spline-based geometry, used in isogeometric analysis, eliminates the mesh-geometry mismatch typical in traditional FEA, leading to improved stress convergence and accuracy in large-deformation and contact problems. Their solver integrates directly with CAD using native spline representations, reducing preprocessing time and enabling consistent simulation fidelity for nonlinear structural applications.
- Rafinex – André Wilmes
ROBUST DESIGNS AND THE ROLE OF VARIABILITY (UQ) WITH RAFINEX
This presentation explored how stochastic optimization incorporates variability in loads, boundary conditions, and material properties directly into the design process, allowing engineers to generate solutions that are not just optimal under ideal conditions but consistently high-performing across a range of real-world uncertainties. The approach was applied to structural components such as engine mounts and gripper arms, demonstrating how robustness can be quantified and validated through integrated simulation workflows.
Industry Applications of Simulation-Driven Design in Architecture, Engineering, and Manufacturing
- Royal HaskoningDHV
GABRIEL GARCIA ON COMPUTATIONAL DESIGN AND MACHINE LEARNING IN MEP SYSTEMS FOR LARGE-SCALE ARCHITECTURE
Garcia presented how performance-based simulation of mechanical, electrical, and plumbing (MEP) systems—such as airflow, Wi-Fi signal propagation, cabling infrastructure, and equipment serviceability—can guide architectural layout decisions early in the design process. By treating MEP as the nervous system of the building rather than an afterthought, his workflow links spatial planning with operational efficiency, influencing not just the façade but the full architectural configuration.
- Andreas Vlahinos
Simulation-Driven Design of Lattice Structures
Vlahinos presented a workflow for evaluating the mechanical behavior of lattice structures under realistic AM-induced imperfections, emphasizing the significance of finally achieving real-time simulation feedback after decades of relying on time-intensive, iterative testing cycles. His approach showcased how engineers can now assess and adjust lattice performance instantly within the design environment, fundamentally changing how optimization and validation are integrated into the design-for-AM process.
- NASA Goddard – Ryan McClelland
FROM TEXT TO SPACESHIP: ADVANCING AI IN AEROSPACE. RYAN MCCLELLAND, NASA
This presentation highlighted how voxel-based topology optimization techniques were used to automate the structural design of lightweight, high-stiffness spaceflight brackets while satisfying mission-specific constraints such as cryogenic stability and vibration resistance. These optimized geometries were directly linked to manufacturing via CNC-compatible toolpath generation, enabling rapid fabrication of fully validated components for NASA missions within a two-week window, integrating simulation at both the part and process levels.
- Ocado Technology – Kiera Exley
Automation-Enabled Design for Additive Robotic Systems
Exley outlined the development of internal tools for parametric design and real-time AM suitability checking for industrial automation. These workflows integrate topology optimization to minimize weight while maintaining performance targets for robotic housings and mounts, using simulation-driven constraints such as deformation limits, part orientation, and support minimization. Input parameters include mechanical load paths and build volume constraints, while outputs include AM-ready geometry and metadata for process planning and cost estimation.
- ARUP – Rick Titulaer
INFORMED DATA-DRIVEN COMPUTATIONAL DESIGN
Titulaer demonstrated data-driven design workflows that integrate urban-scale energy simulations into ARUP’s InForm platform to guide early-stage form generation. The presentation included a stadium design case study where reflection simulation was used to evaluate glare and thermal impact within the architectural envelope, enabling informed geometry adjustments during the conceptual phase.
Upcoming Presentations at CDFAM Amsterdam, July 9-10, 2025
CDFAM Amsterdam will continue to showcase how simulation is not only validating design intent, but actively shaping form, material distribution, and production strategies across domains. The following upcoming presentations highlight the next wave of simulation-driven workflows, tools, and research:
- Mathias Fuchs – McNeel – Rhino Grasshopper II
Isosurface Generation from Simulation Fields in Grasshopper 2
Fuchs will reveal workflows using Grasshopper 2 and TRmesh/TRfem to derive geometry directly from simulated scalar fields, such as temperature and stress, creating performance-driven structures with embedded environmental responsiveness.
- Thomas Rees – ToffeeX
System-Level Optimization through Coupled Multiphysics Simulation
Building on prior work, ToffeeX will present new capabilities in co-optimizing thermal, fluidic, and structural constraints for complex assemblies, showing how early-stage design decisions can be guided by full-system performance predictions.
- Politecnico di Torino – Dario Carbonaro
Computational Design and Optimization of Self-Expandable Cardiovascular Devices
Carbonaro will present computational frameworks for designing vascular stents and TAVs using finite element simulations of implantation, unit-cell level topology optimization, and material tuning through NiTi heat treatment—bridging simulation, material science, and device performance.
- PhysicsX
Engineering at Scale with Large Physics Models
This presentation will outline how pre-trained AI models for aerodynamics, heat transfer, and structural simulation enable engineers to compress months of iterative workflows into minutes, with examples from aerospace and industrial equipment.
- Ansys – Chris Robinson
Design Optimization for Advanced Manufacturing through Forward-Looking Performance Simulation
Robinson will present how simulation of structural, thermal, and multi-physics behavior informs early geometry creation and optimization before manufacturing. Use cases include topology-optimized brackets and heat exchangers, with additional opportunities in RF, NVH, and fluid systems. The workflow couples performance simulation directly to advanced manufacturing methods to achieve optimized, application-specific outcomes.
- Altair – Wesley Essink
Flexible Geometric Modeling and Atypical Simulation Solvers to Streamline Design Optimization
Essink will showcase Altair Inspire’s integrated design optimization workflow, including multi-representation modeling, persistent boundary conditions, and geometry updates directly tied to simulation findings—all executed without switching tools or formats. Applications include the rapid structural optimization of mechanical components, streamlining iterative design cycles across industries.
- Flexcompute – Momchil Minkov + Qiqi Wang
Real-Time Computer-Aided Optimization (CAO): How GPU-Native Simulation Changes the Industry
This talk will demonstrate how GPU-native simulation combined with adjoint and inverse optimization workflows allows for rapid exploration of large design spaces in aerospace and photonics. By integrating real-time, goal-oriented optimization directly into the simulation environment, CAO shifts the design process from reactive validation to predictive synthesis—streamlining development cycles and enabling application-specific performance breakthroughs.
- Cognitive Design Systems – Rhushik MATROJA
Manufacturing Driven Design
Matroja will present how real-time Design for Manufacturing (DfM) checks and process-aware design recommendations enable engineers to generate manufacturable parts across AM, machining, casting, and molding workflows. Proven with partners like Safran and Mitsubishi Electric, the platform integrates simulation and manufacturability analysis into a single software loop—reducing design time while maintaining high performance.
- DTU – Luca Breseghello
Stress-Based Design of Lightweight Horizontal Structures for 3D Concrete Printing
Breseghello will present a simulation-driven workflow integrating stress-based material placement and FEA with robotic 3D concrete printing to develop beams and slabs with significantly improved strength-to-weight ratios. The talk will cover the design and validation of the 3DLightBeam and 3DLightSlab systems, highlighting life-cycle analysis and structural testing for sustainable concrete construction.
- Cornell University – Tiffany Cheng
Bioinspired and Biobased 4D-Printing for Adaptive Building Facades
Cheng will present a computational design workflow integrating material programming, environmental simulation, and 4D printing to create hygromorphic systems for passive climate-responsive facades. Applications span architectural envelopes and self-adjusting bioinspired components across scales.
- KU Leuven – Vanessa Cool
From Structure to Sound: Unlocking the Potential of Vibroacoustic Design
Cool will present a vibroacoustic topology optimization framework for designing meta-structures that balance mass, stiffness, and acoustic performance. Applications span from unit cell and metamaterial design to full-scale structural components, with simulation used to guide design from the early stages and ensure compliance with modern acoustic standards.
Register to Attend CDFAM Amsterdam to Connect with Other Experts in Simulation Driven Design
Simulation is no longer an afterthought or a final check—it is the foundation of modern design workflows across disciplines.
At CDFAM Amsterdam, you’ll engage directly with the researchers pushing the boundaries, the developers building the tools, and the engineers applying these methods to solve complex real-world problems.
Don’t miss the opportunity to meet the people behind these presentations, share your perspective, and build the collaborations that will shape what comes next.
“We learn a lot by just talking to people who are doing completely different things with the same tools.”
— Aaron Porterfield, F=F
