Supported availability
PhysicsMechanical
Mechanical Characterisation Test Simulation
Configuration prototype for compression, tension, diametrical compression, bending and shear tests. Analysis execution is not connected.
Explore capabilityDigital Twin Virtual Laboratory / DTVL
Advanced engineering simulation.
Simplified.
DTVL provides guided, pre-built engineering workflows for advanced simulation, constitutive modelling, model calibration and digital twin development.
Describe your engineering problem. Work through a structured workflow. Interpret the results in the context of your application.
Featured / Figure placeholder
Featured simulation
Reserved for a real technical figure.
Image to be suppliedA featured DTVL simulation image will appear here.
Future figure: geometry, field variable and readable result legend.
Materials & manufacturingSimulation & calibrationData & digital twins
01 / What is DTVL?
An engineering problem.
A structured path to analysis.
Digital Twin Virtual Laboratory brings advanced modelling into guided engineering workflows.
Define the material, geometry and process. Guided setup connects these inputs to computation and engineering results, with assumptions and limitations visible.
For engineers, researchers, manufacturers and technical teams.
02 / How it works
Guided inputs. Connected analysis.
A common workflow structure, from the initial engineering question to simulation and model refinement.
- 01
Engineering inputs
Define material data, geometry, process conditions and objectives.
- 02
Guided model setup
Follow a structured path through model choices and required parameters.
- 03
Simulation
Connect the configured problem to the relevant computational workflow.
- 04
Engineering results
Review predicted behaviour, key outputs and modelling assumptions.
- 05
Calibration / Digital Twin
Compare with measurements and refine the model where appropriate.
03 / Simulation showcase
From process sequence to predicted behaviour.
Simulation figures are reserved below as labelled placeholders. Explore detailed process figures and the formulation sequence on the capabilities page.
S04 / Figure placeholder
ANN-assisted constitutive replacement
Reserved for a real technical figure.
Image to be suppliedComplex 3D shape simulation using ANN replacement of Kp, Gp and Ss in the modified Skorohod–Olevsky model.
04 / Simulation capabilities
Materials, processes and models.
Platform capabilities and planned workflows cover a wide range of powder-processing, thermal, mechanical and data-informed engineering studies. Confirm availability and application suitability with DTVL.
PhysicsMechanical Mechanical Thermal Thermal Electrical Electrical
Badges indicate supported capabilities; active physics and constitutive responses are selected within each model configuration. Coupling availability depends on the selected workflow and material model.
06 / Digital Twin
Connect physical evidence
with numerical prediction.
DTVL uses experimental and process measurements to inform, constrain and update physics-based simulation models. The connection between the physical and numerical systems can serve different purposes depending on the engineering problem.
Experimental feedback and model updating
Experimental observations can be compared with simulation outputs to constrain constitutive parameters, discriminate between admissible material descriptions and progressively reduce model uncertainty.
Different experiments can provide complementary information about volumetric, deviatoric, thermal or other aspects of material behaviour.
Measurement-informed boundary conditions
Real process measurements can also be used directly to prescribe simulation inputs. Measured temperature, force, pressure, displacement or other time-dependent process histories can be applied as loading or boundary conditions so that the numerical model more closely reproduces the conditions experienced by the physical system.
- Physical system
- Measurements
- Model updating / Boundary-condition definition
- Physics-based simulation
- Predicted behaviour
Published University of Leicester research demonstrated digital-twin-style workflows in ceramic sintering, linking measured process information, computational models and predicted deformation. DTVL draws on this experience to connect observations with model predictions and parameter refinement.
Explore constitutive-law and ANN-enabled approaches
The level of digital-twin integration depends on the available measurements, model suitability and application. The current framework should not be presented as automatically providing continuous real-time updating.
DT01 / Figure placeholder
Digital twin calibration loop
Reserved for a real technical figure.
Image to be suppliedExperimental observations are compared with model outputs to refine parameters and improve prediction.
Future figure: observations → model outputs → parameter refinement.
07 / Technology
Physics-based foundations.
Structured computational workflows.
DTVL draws on academic research experience to develop application-specific engineering workflows.
- Finite element simulation
- Physics-based simulation of material and manufacturing processes.
- Advanced constitutive modelling
- Constitutive descriptions incorporating evolving material properties, process history and densification behaviour.
- Automated preprocessing and post-processing
- Guided generation of simulation models and application-specific interpretation of results.
- Model calibration
- Parameter identification using experimental observations and simulation–experiment comparison.
- Machine-learning-assisted workflows
- Neural-network constitutive representations, surrogate modelling and data-assisted parameter identification embedded within FEA workflows.
- Computational infrastructure
- Managed execution of simulation workloads, inputs, outputs and result processing.
T01 / Figure placeholder
Constitutive modelling
Reserved for a real technical figure.
Image to be suppliedMaterial response and constitutive model representation.
Future figure: model schematic or response curve.
T02 / Figure placeholder
Calibration comparison
Reserved for a real technical figure.
Image to be suppliedExperimental observations alongside model predictions.
Future figure: measured-versus-predicted comparison with axes and units.
T03 / Figure placeholder
Machine-learning-assisted workflows
Reserved for a real technical figure.
Image to be suppliedData-assisted methods within a simulation workflow.
Future figure: training data, model and simulation connection.
08 / Research, validation and recognition
Research, validation and recognition.
DTVL is an independently developed platform that builds on research experience. The academic work below was undertaken at the University of Leicester; its results are not claims of DTVL platform validation.
UKRI-funded research
Several modelling and machine-learning concepts presented here draw on postdoctoral research at the University of Leicester through the UKRI Strength in Places Fund – Advanced Ceramics Project no. 82148. Led by Prof. Jingzhe Pan, the university team investigated machine learning, finite element modelling, constitutive behaviour, ceramic processing and digital-twin methods.
Constitutive representations
Neural networks have been investigated to supplement, partially replace or fully represent selected constitutive relationships within FEA. The scope depends on the relationship, available data and validation.
A trainable surrogate could incorporate particle-size distribution, moisture/humidity or batch-to-batch variability. These are potential future inputs requiring appropriate training data and validation, not experimentally validated DTVL capabilities.
Selected publications
Peer-reviewed academic research
-
Physics-based neural network as constitutive law for finite element analysis of sintering
Publisher record -
Machine learning nested in multiphysics finite element analysis: application to flash sintering
Publisher record -
Application of DFEM in sintering distortion analysis of oxide–oxide ceramic matrix composite: digital twin
Publisher record -
A machine learning approach to finite element modeling of sintering deformation using densification data
Publisher record -
Successful Formulation Window for the design of pharmaceutical tablets with required mechanical properties
Polak, P.; Sinka, I.C.; Reynolds, G.K.; Roberts, R.J.
The methodology was developed during doctoral research at the University of Leicester in collaboration with AstraZeneca. This academic provenance is separate from DTVL platform development and company IP claims.
Publisher record
Current research / manuscripts under review
Physics-guided directional backpropagation
Current research investigates material-parameter identification using non-differentiable directional, sign-based backpropagation. Updates follow the direction of the model response rather than the magnitude of a conventional analytical gradient.
- Manuscript under review
Physics-Guided Identification of Material Parameters via Non-differentiable Sign-based Back-propagation in Ceramic Processing
- Manuscript under review
Constitutive Identification in Ceramic Sintering: Reducing Non-Uniqueness through Multi-Rate Dilatometry and Beam Bending
R01 / Figure placeholder
Award-winning poster
Reserved for a real technical figure.
Image to be suppliedResearch recognition: a research poster arising from the Advanced Ceramics programme received second place at a ceramic research event in Birmingham associated with the UKRI Strength in Places Fund programme.
Poster image and event details to be added.
09 / Planned production architecture
Security & Confidentiality
DTVL is being designed around isolation, encryption and short-lived storage of customer engineering data. The planned architecture separates customer models, geometry, material parameters, boundary conditions and simulation results between customer environments and excludes confidential engineering content from routine application logging.
Encrypted
The production design calls for customer data to be protected in transit and at rest using secure transport and modern authenticated encryption.
Isolated
Customer storage, simulation jobs and customer-specific models are intended to be separated through server-side access controls and isolated compute workflows.
Short-lived
Simulation files and results are intended to be retained for a short predefined retention period before automatic deletion. The production design also calls for destruction of per-job encryption keys for cryptographic erasure.
Minimal exposure
The planned logging policy excludes geometry, material parameters, boundary conditions and simulation results from routine operational logs, retaining only the minimum metadata required for system operation and security monitoring.
Customer-controlled result protection
The planned architecture supports storing simulation results with a dedicated per-job encryption key. A future production workflow may also support customer-specific encrypted download packages, including public-key encryption where appropriate, so that only the customer holding the corresponding private key can decrypt the exported data.
Customer confidentiality
The intended production policy keeps customer geometry, material definitions, constitutive parameters, loading histories, boundary conditions, simulation results and customer-specific models confidential to that customer. They will not be exposed to other customers or reused for other customer workflows without explicit agreement.
Administrative access to production customer data will be restricted, role-controlled and auditable. DTVL personnel will not routinely inspect customer simulation content.
- Secure upload
- Isolated customer workspace
- Encrypted simulation storage
- Isolated FEA execution
- Encrypted results
- Customer download
- Automatic deletion + key destruction
Security architecture under development
Final production controls, retention guarantees and customer-data handling policies will be published before the service accepts confidential customer data.
10 / Customer Portal
Your entry point to DTVL.
Access the DTVL web application, or enquire about workflow suitability and customer access.
11 / About
Digital Twin
Virtual Laboratory.
DTVL focuses on guided simulation, materials modelling and digital twin development for engineering and scientific applications.
DTVL is an independently developed platform. References to university research describe the academic origin of underlying methods and do not imply university ownership, endorsement, equity or operational involvement in DTVL.
12 / Contact
Start with your
engineering question.
Define the material, process or component you want to understand.
A useful application brief includes your intended outcome, available measurements, process conditions and any existing modelling work.
- General enquiries
- info@digtwinlab.com
- Technical support
- support@digtwinlab.com
- Engineering enquiries
- peter.polak@digtwinlab.com
- Billing
- billing@digtwinlab.com
- Security / responsible disclosure
- security@digtwinlab.com
Additional company, telephone and registered-office details will be added as they become available.