Multivariate measures of in-process wood pulp composition and morphology: smart sensors for real-time process control and fibre product optimization

来自加拿大北方针叶木的漂白硫酸盐纸浆在纤维长度、强度、细度和均匀性方面处于世界领先地位。然而，造纸技术的进步缩小了NBSK纸浆的优势，增加了加拿大供应商的性价比压力。为了保持竞争力，加拿大纸浆行业必须创新，开发新的工艺和过程控制策略，以生产质量更高、成本更低的纸浆，同时减少能源消耗和对环境的影响。纸浆的化学和材料特性决定了其产生的纤维网络的物理特性，在充分的过程监控和干预下，有机会制造符合纸浆和最终用途纸张规格的纸浆。然而，为了使过程控制取得成功，生产商必须不断地进行测量，并迅速做出反应。纸浆流中未被识别的变异会产生不符合规格的产品，从而浪费资源，而持续的过程监控可以确保最佳的输出质量，这是由当时的记录证明的。在这里，我们提出的研究将为生产线带来新的激光光谱分析工具。我们的方法将结合小型化仪器、集中采样协议、精细化的多变量校准和集成分析，以实现针对光纤网络产品物理和机械性能的实时过程控制。面向模板的遗传算法(TOGA)特征选择将在纸浆流拉曼光谱中连接化学标记，以监控与最终用途机械规格的一致性，如抗张强度、耐破度、撕裂度和纸浆粘度。TOGA功能将发出与生物特征参数相关的化学变化信号，要求进行工艺补救和/或最终产品重新分配。离线测量将把化学信息与纤维系统形态的亚微米空间测量融合在一起，以增加对网络强度和柔软度的基本了解。最后，实地研究将应用这些相同的化学和形态分类工具来绘制正在使用的电子材料纤维系统的故障进展图。

Bleached kraft pulps derived from Canadian northern softwood lead the world in terms of fibre length, strength, fineness and uniformity. However, progress in paper manufacturing technology has narrowed the advantage of NBSK pulps, increasing price-performance pressure on Canadian suppliers. To remain competitive, the Canadian pulp industry must innovate, developing new processes and process-control strategies that yield pulps of higher quality and lower cost, with reduced use of energy and impact on the environment. The chemical and material properties of a pulp determine the physical properties of the fibre network it produces, and with adequate process monitoring and intervention, the opportunity exists for manufacturing to target pulp and end-use paper specifications. However, for process control to succeed, producers must measure continuously and react quickly. Unrecognized variation in a pulp stream wastes resources by yielding a product that fails to meet specifications, while continuous process monitoring can ensure optimum output quality, certified by a contemporaneous record. Here we propose research that will bring new, laser-spectroscopic analytical tools to the production line. Our approach will combine miniaturized instrumentation, focussed sampling protocols, refined multivariate calibration and integrated analysis to enable real-time process control targeted to physical and mechanical properties of fibre network products. Template oriented genetic algorithm (TOGA) feature selection will link chemical markers in pulp-stream Raman spectra to monitor conformance with such end-use mechanical specifications as tensile strength, burst, tear, and pulp viscosity. TOGA features will signal chemical changes that correlate with biometric parameters, calling for process remedies and/or end-product reallocation. Off-line measurements will fuse chemical information with sub-micron spatial measures of fibre system morphology to increase fundamental understanding of network strength and softness. Finally, field research will apply these same chemical and morphological classification tools to map the progress of in-use electronic-materials fibre systems to failure.