Laboratory experiments investigating syn-eruptive crystallization are fundamental for interpreting crystal and vesicle textures in pyroclasts. Previous experiments have advanced our understanding by varying decompression and cooling pathways, volatile components, and melt composition. However, they have largely failed to produce the high crystal number densities seen in many cryptodome and dome samples. In this study, we approach the problem by exploring non-linear decompression pathways. We present two series of experiments: (1) decompression from low initial starting pressure and (2) a compression-and-release step after the initial decompression. The purpose of each series was to simulate (1) decompression of magma that stalls during ascent and (2) pressure cycling that occurs in non-erupted magma that is decompressed in the conduit during episodic explosive activity. The experiments were carried out on a synthetic rhyodacite (SiO2 = 69 wt. %) held initially at 50 MPa and 885oC then decompressed at rates of 0.026-0.05 MPa s-1 to 10 MPa. A subset of experiments was then subjected to a compression step to 110 MPa followed by near-instantaneous release back to 10 MPa. A substantial volume fraction of dendritic microlites formed during the initial hold at 50 MPa; additional crystallization during subsequent decompression to ≥10 MPa was minimal, as evidenced by only small increases in crystallinity and comparable crystal number densities. Samples that underwent recompression followed by a second decompression showed no increase in crystal volume fraction but did show extensive disruption of the initial dendritic, box-work microlite structures that produced high number densities of small individual crystals. The disruption was driven by a combination of rapid vesiculation, expansion and resulting shear along the capsule walls. From these results, we suggest that high crystal number densities may be a signature of rapid deformation occurring after magma stalling in the subsurface, perhaps related to pressure cycling and accompanying rapid changes in vesicularity during repeated small and shallow-sourced explosions. We compare our experiments to pyroclasts from shallow intrusions that preceded the May 18 1980 of Mount St Helens (erupted both before and during the eruption) to illustrate the significance of our results.

中文翻译：

---

机械损伤导致的高晶体数密度

研究协同增生结晶的实验室实验是解释火山碎裂中晶体和囊泡质地的基础。以前的实验通过改变减压和冷却途径，挥发性成分和熔体组成，进一步提高了我们的理解。但是，它们在很大程度上无法产生许多隐球菌和圆顶样品中所见的高晶体数密度。在这项研究中，我们通过探索非线性减压途径来解决这个问题。我们提供了两个系列的实验：（1）从较低的初始启动压力开始减压，以及（2）在初始减压之后进行压缩和释放步骤。每个系列的目的是模拟（1）在上升过程中失速的岩浆减压和（2）在爆发性爆炸活动期间在导管中减压的非喷出岩浆中发生的压力循环。实验是在最初保持在50 MPa和885oC，然后以0.026-0.05 MPa s-1到10 MPa的速率减压的合成流纹岩（SiO2 = 69 wt。％）上进行的。然后将实验的子集进行压缩至110 MPa的步骤，然后立即释放回10 MPa。初始保持在50 MPa时形成的大量树枝状微晶；在随后的减压至≥10 MPa的过程中，额外的结晶是极少的，这可以通过结晶度和相当数量的晶体密度的小幅增加来证明。经过再压缩再进行第二次减压的样品未显示晶体体积分数的增加，但确实显示了最初的树枝状，盒式微晶结构的广泛破坏，该结构产生了高密度的单个小晶体。破裂是由快速囊泡形成，扩张和沿囊壁产生的剪切力共同作用引起的。根据这些结果，我们认为高晶体数密度可能是岩浆在地下失速后发生快速变形的标志，这可能与压力循环以及在反复的小型和浅源爆炸中伴随的囊泡快速变化有关。