The sustained transportation of particles in a suspension commonly results in particle attrition leading to grain size reduction and shape modification. Particle attrition is a well-studied phenomenon that has mainly focussed on sediments produced in aeolian or fluvial environments. Here, we present analogue experiments designed to explore processes of attrition in the kimberlite system; we focus on olivine as it is the most abundant constituent of kimberlite. The attrition experiments on olivine use separate experimental set-ups to approximate two natural environments relevant to kimberlites. Tumbling mill experiments feature a low energy system supporting near continual particle-particle contact and are relevant to re-sedimentation and dispersal processes. Experiments performed in a fluidized particle bed constitute a substantially higher energy environment pertinent to kimberlite ascent and eruption. The run-products of each experiment are analysed for grain size reduction and shape modification and these data are used to elucidate the rates and extents of olivine attrition as a function of time and energy. Lastly, we model the two experimental datasets with an empirical rate equation that describes the production of daughter products (fines) with time. Both datasets approach a fines production limit, or plateau, at long particle residence times; the fluidized system is much more efficient producing a substantially higher fines content and reaches the plateau faster. Our experimental results and models provide a way to forensically examine a wide range of processes relevant to kimberlite on the basis of olivine size and shape properties.

中文翻译：

---

金伯利岩系中的磨损

悬浮液中颗粒的持续运输通常会导致颗粒磨损，从而导致颗粒尺寸减小和形状改变。颗粒磨损是一种经过充分研究的现象，主要集中在风成或河流环境中产生的沉积物上。在这里，我们展示了旨在探索金伯利岩系统磨损过程的模拟实验；我们专注于橄榄石，因为它是金伯利岩中含量最丰富的成分。橄榄石的磨损实验使用单独的实验装置来近似与金伯利岩相关的两个自然环境。滚筒磨机实验具有低能量系统，支持近乎连续的颗粒-颗粒接触，并且与再沉淀和分散过程相关。在流化颗粒床中进行的实验构成了与金伯利岩上升和喷发有关的高得多的能量环境。分析每个实验的运行产物的晶粒尺寸减小和形状修改，这些数据用于阐明作为时间和能量函数的橄榄石磨损的速率和程度。最后，我们使用经验速率方程对两个实验数据集进行建模，该方程描述了子产品（细粉）随时间的产生。在长颗粒停留时间下，两个数据集都接近细粉生产限制或平台；流化系统效率更高，产生更高的细粉含量并更快地到达平台。