Enhancing the pumpability of solid-waste-based cement mortar: understanding the coupled effects of ultrasonication on rheological properties via support vector machine modeling.

Solid-waste-based cement mortar (SWCM), prepared with ultrafine tailings and cement as the main raw materials, is widely used in mining backfilling and construction engineering. However, SWCM tends to form dense flocculation networks, inducing ultra-high yield stress and plastic viscosity that severely limit its pumpability. The objective of this study was to quantify the coupled effects of ultrasonic frequency (20-40 kHz), power (25-100 W), solid mass fraction (70%-78%), and binder-to-fine aggregate mass ratio (1:6-1:10) on the plastic viscosity and yield stress of SWCM and to establish a support vector machine (SVM)-based prediction framework for pumpability-oriented parameter optimization. The results showed that ultrasonic treatment effectively reduced both the plastic viscosity and yield stress of SWCM. Under the optimal parameters tested here (40 kHz, 100 W), these two parameters decreased by averages of 39.33% and 34.27%, respectively, relative to untreated samples. Moreover, the established SVM model showed good predictive performance, with correlation coefficients of 0.9013 and 0.9322 for plastic viscosity and yield stress, respectively. This work clarifies the rheological response of SWCM to representative ultrasonic conditions and provides a data-driven reference for mix design and pumping-process optimization of solid-waste-based cementitious materials.