Newsletter October 2026

Newsletter October 2026

Dear Sir or Madam,

 

In this newsletter, you can look forward to exciting topics and up-to-date insights:

 

Hundreds of Simulation Scenarios with Just a Few Prompts

Building models, configuring variants, running simulations and evaluating results: AI agents can accelerate many of these demanding tasks or even perform them automatically. At TLK, we are actively driving this development forward with TMODE (TLK Modelica Development Environment). Our development environment, built on Dymola and OpenModelica, brings together modeling, simulation and generative AI within Visual Studio Code.

Instead of spending significant time manually building and parameterizing simulation models of thermal systems, you can delegate this work to AI agents with the help of TMODE. Our platform enables agents to create, modify, connect and simulate models directly through standardized tool interfaces. Integrated model checks provide immediate feedback after every modification, allowing errors in syntax, structure and parameterization to be detected and corrected at an early stage. Meanwhile, you can focus entirely on the engineering assessment and interpretation of the results.

 

For agents to generate not only correct code but also technically meaningful models, they can access a curated knowledge base developed by us. This knowledge base contains thermodynamic expertise, model information (e.g., from the TIL library) and application guidelines. This knowledge base is currently still under development and continuous refinement. To this end, we are investigating approaches such as Retrieval-Augmented Generation (RAG) and structured Markdown databases such as LLM-Wiki. A distinctive feature of TMODE is therefore the combination of the capabilities of generative AI with our proven TIL components and our documented domain expertise.

 

In addition, AI agents in TMODE can make effective use of other TLK software solutions:

 

In combination with MoBA, our tool for automated simulation studies, AI agents configure extensive parameter studies, coordinate simulation runs and evaluate results – while the evaluation criteria can, of course, be defined by the user. Through an MCP connection to DaVE, agents can also generate the desired visualizations. This means you no longer need to spend time searching for the most suitable way to present results and can instead focus directly on their technical interpretation.

 

Although the AI agents in TMODE take over many tasks, you as the user remain in full control at all times. The generated Modelica code can be reviewed and edited either in the diagram view or directly in the source code, supported by syntax highlighting and linting in Visual Studio Code. In addition, AI agents can access only those directories that you explicitly make available to them.

 

The opportunities opened up by AI are driving a paradigm shift in the way we work and we want to shape this transformation together with you. While we remain mindful of the associated risks, we also see countless new possibilities emerging. Some of them are already within reach, while others are opportunities we cannot even envision today.

 

TMODE is currently a prototype that we use internally in our day-to-day development work and continuously enhance. If you are interested in AI-assisted model development, would like to discuss the topic with us or want to contribute your own workflows and use cases, please feel free to contact us directly.

Register Now: Software Webinar Featuring TIL Toolkit and Compare Time Series

The new releases of our in-house software products are now available. To introduce the latest features and developments, we invite you to join our next software webinar at the end of October: on October 28 in German and on October 29 in English. During this webinar, we will present our latest developments in the field of AI and showcase new features and enhancements in the TIL Suite, DaVE, and MoBA Automation.

In addition, we would like to introduce our new TIL Toolkit, a collection of practical tools designed to simplify your work with TIL Suite in Dymola. Its user-friendly features support you in recurring modeling, analysis and visualization tasks in the field of thermal simulations, helping you streamline workflows, save time and work more efficiently.

A special focus of this webinar will be on the Compare Time Series tool. It features automated signal matching, advanced comparison capabilities, and seamless integration with DaVE. With this tool, development teams can efficiently identify, analyze and document differences – whether between simulation and measurement data or between different model versions. As a result, validation, regression testing and root-cause analysis are simplified and accelerated.

 

Our software experts will be available throughout the webinar to answer your questions and discuss your specific use cases.

If you are interested, you can register free of charge via the button on our webinar website. We look forward to your participation and to an engaging technical exchange!

The TIL Family Is Growing: Steady-State Simulations with TIL ST

With the development of TIL ST, we are completing the model portfolio of the TIL Suite: In addition to libraries for dynamic simulations (TIL) and highly dynamic simulations (TIL HD), TIL ST will soon provide a library based on native steady-state modeling approaches. As a result, the TIL Suite will cover all phases of a typical thermal system development workflow: from the conceptual design and evaluation of system topologies, through the sizing of individual components, to the development of dynamic operating and control strategies.

 

TIL ST offers the following features:

 

  • Direct specification of operating conditions such as superheating, subcooling and filling levels in receivers
  • Calculation of geometric or operating parameters
  • Fast and easy conversion between steady-state and dynamic TIL models
  • Graphical debugging of models and the solution process

The different TIL variants can be applied across the various phases of a typical development process as follows:

 

  • Optimization through large-scale simulation studies: TIL ST with NTU-based heat exchanger models
  • Component sizing: TIL ST with geometry-based heat exchanger models
  • Development of operating and control strategies: TIL for dynamic simulations and TIL HD for highly dynamic processes

 

The first release of TIL ST is planned for spring 2027. For further information or if you are interested in evaluating a pre-release version, please contact:

Efficient and Future-Proof Data Center Cooling with TIL Suite

With the rapid growth of AI applications and increasing requirements for energy efficiency and waste heat utilization driven by regulations such as the German Energy Efficiency Act, data center cooling is becoming increasingly important. Modern AI servers achieve significantly higher power densities than conventional IT systems and consequently generate large amounts of heat that must be removed reliably. In addition to established air-cooling systems, innovative solutions such as liquid cooling, direct two-phase cooling, and immersion cooling are gaining popularity. As a result, selecting and designing the optimal cooling concept is becoming increasingly complex.

 

Simulations enable the reliable evaluation of different concepts at an early stage of the design process, support component sizing and help optimize control strategies. Key performance indicators such as energy consumption, Power Usage Effectiveness (PUE) and waste heat recovery potential can be analyzed and assessed long before implementation.

We have extensively investigated the cooling of modern data centers and developed and analyzed representative system models using our TIL Suite. The current release includes a complete example system that can serve as a starting point for your own studies and the development of customized models. As part of our analyses, we performed a detailed assessment of control concepts using the Relative Gain Array (RGA) method, compared pure free-cooling and evaporative hybrid cooling concepts and further improved the simulation performance and robustness of the models.

 

Would you like to perform similar analyses for your own systems? We would be happy to support you or provide suitable reference models as a starting point. For further information, please contact us directly:

TLK On Site: TLK-Thermo and TLK Energy at Chillventa

Autumn is just around the corner and with it comes one of the most important industry events for refrigeration, air conditioning and heat pump technology: Chillventa 2026 in Nuremberg. We are very pleased to be exhibiting this year together with TLK Energy.

 

Visit us from October 13 to 15, 2026, at booth 4-509 in Hall 4. Take the opportunity to discuss current challenges and developments in heat pump and refrigeration technology with our experts. Whether you are interested in customized test bench development, dynamic system simulation, optimized control strategies or the design of complex thermal systems: We look forward to exchanging ideas and discussing your specific applications and requirements in person.

We will demonstrate how our testing capabilities and software tools can support and accelerate your development processes. In addition, we will showcase an exciting demonstrator from our laboratory and have our TLK-Thermo-Meters available for you to explore.

Stop by, bring your challenges with you and take advantage of this opportunity for in-depth technical discussions and personalized advice. We look forward to your visit and to speaking with you in person in Nuremberg!
For the latest updates on our activities, you are also welcome to follow us on LinkedIn.

 

With innovative ideas, exciting events and our many years of expertise, we aim to foster professional exchange with you. We look forward to insightful conversations, fresh perspectives and collaborative ideas for current and future challenges.

 

Kind regards,
Willi Tegethoff und Niko Lemke