The Challenge
- There was a need for numerical simulation when a request came in from a company that manufactures high-end server computers.
- There was a serious issue regarding the cooling of the CPU, which needed a tailored solution to the problem.
- In order to obtain the best design of a cooling module, there were many parameters that needed to be optimized.
- This pushed the team at Forwiz to start conducting CAE simulations, leading them to SimScale’s cloud solution .
- The results from the simulation gave the team insights into the parameters to be optimized, ultimately facilitating discussions with future clients on proposed concepts.
Thermal Simulation Conclusions
- The simulations performed gave the team reasonable results, and matched well with the actual experiments performed later on.
- There were more than 100 simulation runs and optimization variations performed using SimScale.
- The requirement was about 10 to 20 core hours to optimize each part, and between 80 to 150 core hours to run studies on entire assemblies.
About Forwiz System
Forwiz System Co., Ltd offers IT services mainly focusing in the field of system integration, network, and software development. It has a recently launched division running thermal management projects such as cooling of high-end server computers and thermal management of batteries used in electric vehicles mainly utilizing heat pipes. There are about 70 employees working in the company, and 5 people working in the thermal division.
Thermal Management of High-End Server Computers
There was a need for numerical simulation when a request came in from a company that manufactures high-end server computers. There was a serious issue regarding the cooling of the CPU, which needed a tailored solution to the problem. In order to obtain the best design of a cooling module, there were many parameters that needed to be optimized.
This pushed the team at Forwiz to start conducting CAE simulations, leading them to SimScale’s cloud solution. The results from the simulation gave the team insights into the parameters to be optimized, ultimately facilitating discussions with future clients on proposed concepts.
Cloud-Based Conjugate Heat Transfer
The choice to go with SimScale was largely influenced by its cloud-based nature. This meant it didn’t require additional computational resources and physical space. Considering the size of the projects, the cost was a great match for the team at Forwiz. Additionally, the many on-demand tutorials helped them get started immediately with SimScale’s simulation software, and support provided assistance whenever necessary.
The team first began with a conceptual design, and then divided the design into separate parts like a heat pipe, connector, and heat sink. Next, the team began to define parameters (i.e, the distance between fins in the heat sink and magnitude of airflow) that required optimization.
The team then started simulating, varying the parameter to see the tendency. After the optimization process of each part, the Forwiz research team evaluated the overall performance of the assembled product and whether it matched the final requirement. During this process, conjugate heat transfer analysis type which included heat transfer taking place between solid and fluid parts was chiefly used, in addition to gathering insights on airflow patterns.
Seongsu Park
Researcher
A few highlights from my experience with SimScale are that UX/UI is very intuitive and easy to use, and as SimScale is running in the cloud we can work with the simulation and easily check/share the result no matter where we are. Since we are able to provide the clients with realistic data incorporating reasonable numbers generated with the relevant simulation, it has become much easier to discuss the problem to solve in the initial stage, thus expanding business opportunities.
Accurate & Fast Results
The simulations performed gave the team reasonable results, and matched well with the actual experiments performed later on. There were more than 100 simulation runs and optimization variations performed using SimScale. The requirement was about 10 to 20 core hours to optimize each part, and between 80 to 150 core hours to run studies on entire assemblies.
Continued Collaboration on Thermal Management Projects
For a detailed overview of Forwiz System’s simulation setup and results, sign up to download our on-demand webinar on CPU cooling with cloud-based simulation
