Error Modeling and Error Control Study of PA/Pine Wood Biomass Composites

Area of Science:

• Materials Science
• Additive Manufacturing
• Mechanical Engineering

Background:

• Laser sintering (LS) is a key additive manufacturing technology, but accuracy issues limit its civilian applications.
• Polyamide (PA) parts fabricated via LS often suffer from dimensional errors.
• Biomass materials show potential as fillers to enhance the printing accuracy of LS parts.

Purpose of the Study:

• To analyze the material properties of PA/pine biomass composites.
• To investigate error control methods for LS-fabricated parts using PA/biomass feedstock.
• To develop and compare theoretical and data-driven models for predicting and compensating dimensional errors.

Main Methods:

• Established a theoretical mathematical model incorporating material properties, process parameters, and equipment factors.
• Developed a data-driven model using BP neural networks based on experimental data.
• Examined the predictive capabilities and compensation effects of both models.

Main Results:

• A nylon/pine wood biomass composite with 3 wt% pine content yielded parts with 20 MPa tensile strength and a 10 °C sintering preheating window.
• Both theoretical (62-73%) and data-driven (81-91%) models predicted dimensional deviations.
• Compensation reduced overall dimensional deviation from 1.61-3.49% to 0.41-0.50%.

Conclusions:

• PA/biomass composites offer a viable solution for improving LS part accuracy.
• The data-driven model significantly outperforms the theoretical model in predicting dimensional errors.
• Error compensation strategies effectively enhance the precision grade of LS-fabricated parts.