Tzu-Chien Kuo

Mr. Tzu-Chien Kuo

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 August 2024 in The International Journal of Advanced Manufacturing Technology
Verified
Journal Article
Closed access
PDF file
CWChen-Zhe Wu
Chen-Zhe Wu
WHWen-Hsin Hsieh
Wen-Hsin Hsieh
TKTzu-Chien Kuo
Tzu-Chien Kuo
YZYu-Hao Zhang
Yu-Hao Zhang
WFWei-John Fong
Wei-John Fong
JZJyun-Rong Zhuang
Jyun-Rong Zhuang

21 March 2023 in Machines
Verified
Journal Article
Open access
PDF file
TKTzu-Chien Kuo
Tzu-Chien Kuo
AYAn-Shik Yang
An-Shik Yang
YHYih-Chyun Hwang
Yih-Chyun Hwang
WHWen-Hsin Hsieh
Wen-Hsin Hsieh
In the field of precision machining, temperature fluctuation tends to cause the most significant machining errors. In particular, heat, which is generated in the nut of the ball screw feed system during movement, can deform the screw shaft significantly. In order to calculate and evaluate the thermal deformation of the ball screw shaft, the rate of the heat transfer from the nut to the screw shaft must be known. This rate can be calculated by subtracting the heat transfer rate to the nut raceway from the heat generation rate of the nut. Hence, it is necessary to calculate the heat flux from the nut to the nut raceway. This paper introduces a novel method to calculate the heat flux from the nut to the nut raceway. The new approach also enables calculations for different operating conditions. Furthermore, an experimental setup is established to measure the temperature increase, from 0 to 180 s after the nut starts moving, for various operating conditions. It is then theoretically shown that the 0–180 s temperature increase/heat flux curves for the nut are “universal”, i.e., the curve remains unchanged for the different operating conditions. Subsequently, a thermal model using the finite element method (FEM) is developed to simulate the nut temperature increase over time, which is then compared with the experimental data. As a result, it becomes possible to determine the heat flux from the nut to the nut raceway and calculate the 0–180 s temperature increase/heat flux curve (ΔT˜0~180s,Training Data) for the training group. Finally, the heat flux from the nut to the nut raceway is calculated for ten different operating conditions in the test group using the 0–180 s temperature increase/heat flux curve of the training group (ΔT˜0~180s,Training Data). The corresponding temperature curves are then calculated by inputting the values of the heat fluxes into the FEM model. The highest root mean square error (RMSE) between the calculated and experimentally measured temperature increase was 0.16 °C for Test 7 (the error was 10.7%). This result indicates that the new method is valid and feasible for calculating the heat flux from a ball screw nut to the nut raceway.

9 March 2026 in Machines
Verified
Journal Article
Open access
PDF file
CCChia-Chin Chuang
Chia-Chin Chuang
ZCZheng-Wei Lin Chi
Zheng-Wei Lin Chi
TKTzu-Chien Kuo
Tzu-Chien Kuo
CCCHE JUI CHANG
CHE JUI CHANG
WHWen-Hsin Hsieh
Wen-Hsin Hsieh
Structural aging and environmental changes associated with long-term operation can substantially modify the thermal behavior of machine tools, diminishing the accuracy of existing thermal compensation models. Traditional neural network approaches typically necessitate time-consuming and inefficient retraining from scratch for practical adaptation. To address this limitation, this study proposes a parameter-based transfer learning technique to enhance model adaptability under evolving machine tool operating conditions. The method establishes a composite fine-tuning architecture by adding hidden layers and selectively freezing neural network parameters, enabling the rapid adaptation of the pretrained model to new thermal characteristics using limited data. A full-factorial experimental design identified the optimal configuration, comprising (i) structural expansion via an LSTM layer inserted after the hidden layers; (ii) a strategy freezing parameters in all layers; and (iii) training under the selected optimal condition (C9), which reflects machine tool characteristics and environmental temperature variations. The baseline model achieved an RMSE of 3.88 µm. Traditional retraining using the complete dataset and retraining only on C9 yielded RMSE values of 3.21 and 3.84 µm, respectively. In contrast, the optimized transfer learning model trained on C9 achieved an RMSE of 3.47 µm. Experimental results demonstrate that the proposed strategy converges with limited data, reducing the number of datasets from 18 to nine and significantly shortening training time from 18 h 20 min to 30 s. This approach offers an effective solution for sustainable model maintenance and expedited industrial deployment.