Abstract
Manufacturing imperfections are an inherent aspect of the production process, affecting the reliability and performance of all engineered components, including actuators. This work investigates the effects of manufacturing uncertainty in coil assembly. Given the manufacturing tolerances of geometrical and material properties of wires and coils, the coil’s electrical response is studied from a probabilistic perspective. While many researchers acknowledge these uncertainties, they often do not quantify their effects on the system’s response. We propose the use of copulas to model nonlinear relationships and quantify the probabilistic dependence between design variables (i.e., wire and coil’s properties) and their effects on electrical responses. Additionally, tolerances of the design variables are determined through unstructured expert elicitation. The results show the impact of the thickness of the wire’s insulation in the performance of the coil. Furthermore, we establish a design space that is studied probabilistically, allowing for probabilistic design optimization.
| Original language | English |
|---|---|
| Number of pages | 16 |
| Publication status | Published - 2025 |
| Event | 6th International Conference on Uncertainty Quantification in Computational Science and Engineering, UNCECOMP 2025 - Rodos Palace International Convention Center, Rhodes, Greece Duration: 15 Jun 2025 → 18 Jun 2025 Conference number: 6 https://2025.uncecomp.org/ |
Conference
| Conference | 6th International Conference on Uncertainty Quantification in Computational Science and Engineering, UNCECOMP 2025 |
|---|---|
| Abbreviated title | UNCECOMP 2025 |
| Country/Territory | Greece |
| City | Rhodes |
| Period | 15/06/25 → 18/06/25 |
| Internet address |
Keywords
- Uncertainty Quantification
- Copula
- Coreless Electric Motor
- Coils
- Multi-physics model
Fingerprint
Dive into the research topics of 'Uncertainty Quantification of Wire Tolerances in Coil Assembly'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver