METHODS AND EXPERIMENTAL TECHNIQUES

3.1 Introduction

Seven samples o f porous quartz were shock loaded with different loading paths, using a flat plate accelerator. This chapter falls into four parts to describe the methods and experimental techniques employed in these experiments. The first concerns the porous quartz sample material and the determination o f its properties. The second part describes the design o f the sample assembly that contained the sample material during the experiment and a description o f the single stage powder gun used to accelerate the flyer plate into the sample assembly to produce the shock wave. The final two parts describe the post recovery preparation o f the sample material and then the analytical techniques used to assess the extent o f shock metamorphism in the recovered samples.

3.2.1 Selection of sample material

Shock metamorphic effects in quartz are the most extensively investigated and published o f all minerals. Quartz also exhibits the greatest range o f shock features over shock

conditions o f the minerals that have been studied ("see Chapter J), allowing a greater resolution

in classifying the extent o f shock metamorphism and attempting to calibrate observed shock conditions.

Porous quartz, i.e. a SiO] powder, was a preferable sample target material to a ‘solid’ single crystal o f quartz for two reasons; firstly, it magnifies the loading path effects; secondly, it provides numerous crystallographic orientations for the shock wave to pass through. The magnification o f the loading path effects is due to the difference in loading path compared to a

chemically non-porous identical material, illustrated graphically below {figures 3.1 and 3.2).

The greater compressibility o f porous material results in greater net internal energy increases compared to the non-porous equivalent (Kieffer, 1971, Batsanov, 1994). The waste heat is distributed more heterogeneously on a granular scale, in a porous material, with the greatest amount o f heating occurring in pore spaces and at grain boundaries as these spaces are rapidly compressed shut or filled with melt (Kieffer, 1976). The higher net internal energy increases o f porous quartz are desirable to accentuate any differences in the extent o f shock metamorphic effects due to loading path.

The porous sample, composed o f many randomly orientated quartz grains allows the shock front to pass through the maximum number o f crystallographic orientations in a single experimental sample. Thus, the orientation o f the sample with respect to the shock front is no longer a variable that will influence the formation o f shock metamorphic effects over the entire sample. In addition, a large number o f different crystallographic orientations o f the quartz sample can be observed in a single TEM sample or optical microscope thin section slide.

20 18 16 S I'* & I2 2 10

I «

Figure 3.1, P - V plot showing net internal energy increase (shaded yellow) for a porous quartz (density 1.43 cc/g) loaded to ~I6 GPa. (Release adiabat is assumed to approximate the Hugoniot curve for quartz).

Specific Volume (cc/g) 20 18 16

i

j :

Figure 3.2, P - V plot showing net internal energy increase (shaded yellow) for ‘nonporous’, single crystal quartz shocked to ~I6 GPa. (Release adiabat is assumed to approximate the Hugoniot curve for quartz).

Specific Volume (cc/g)

3.2.2 O rigin and description of qu artz sam ple m aterial

The sam ple material is a fine-grained silica sand extracted from a batch o f “ Redhill 110” washed and graded high silica sand from the Cretaceous lower greensand. It was quarried in Redhill, Surrey in the U.K. by, and purchased from, Hepworth M inerals and Chem icals Ltd. based in C heshire, U.K. The grain size o f the quartz sand used the sam ple material was dependent on a num ber o f factors. In shock experim ents the sample material ideally has a large diam eter to thickness ratio. In these experim ents a ratio o f - 1 0 : 1 was used resulting in the cylindrical sam ple having a 10 mm diam eter with a I mm thickness. To minimize interface effects sam ples need to be 10 or more grains in thickness therefore the grain size was kept below 100pm. To retain ease o f examination under the petrological m icroscope grain size was chosen to remain greater than the standard thin section thickness o f 30pm . Standard Endecott brass sedim entalogical sieves were used to grade the material according to grain size. The sieves in standard sedim entological mesh sizes that fitted the above criteria were 63 pm and 75 pm, hence the porous quartz sample material had a grain size ranging from 63 pm to 75 pm.

3.2.3 Pre-shock p reparation of sam ple m aterial

Grains between 63 and 75 pm were separated from the bulk o f the sand by sieving. The standard sedim entological Endecott brass sieves were stacked with a collecting pan at the base.

Above the collecting pan was a 63 pm mesh sieve then a 73pm sieve, between which sand o f the desired grain size accumulated after 20 minutes o f agitation.

Once a quantity o f the correct sized sand had been separated, it was washed using acid to dissolve away any trace amounts o f non-siliceous minerals and impurities to leave pure SiO] sand (Hough, pers.com., 1998). First, the sand was washed in concentrated Hydrochloric Acid (HCl) to remove any glauconite or carbonates that might be present. After rinsing with water, chromic acid was used to dissolve away any organic matter. For the treatment with HCl, the sand was placed in acid-proof plastic containers and twice as much HCl acid to sand was added. The acid and sand were thoroughly mixed. When the sand had settled out the excess acid was decanted off. This procedure was repeated until the acid remained clear indicating the glauconite and carbonate (which colours the acid green) had dissolved away. Then the sand was repeatedly washed through with distilled water and the pH tested until the water retained a neutral pH.

The sand was transferred to glass receptacles for the removal o f any organics with chromic acid. The chromic acid was produced by combining sulphuric acid with dichromate. The chromic acid was added to the sand, stirred and left for eighteen hours for the organic matter to dissolve away, and for the sand to settle to the bottom o f the glass container. The chromic acid was decanted o ff the sand and water was added to the acid to commence the washing procedure. Rapidly adding the largest possible quantity o f cold, distilled water to the acid saturated sand, counteracted the potentially explosive, heating reaction initiated by adding water to acid. The remaining chromic acid was washed out o f the sand by adding water allowing the sand to settle and decanting the mixture o f distilled water and chromic acid, until the sand was free from the acid. The pure quartz sand was dried on a foiled covered tray in an oven kept at a constant 30°C.

3.2.4 Physical properties of porous quartz sample material

The physical properties, packing density, porosity, chemical composition, grain size and shape o f the porous quartz sample material were constrained by measurement and calculation. From the density o f the porous quartz, a suitable Hugoniot curve can be defined.

Table 3.1, Summary o f porous quartz properties

Grain Size Porosity Density Grain shape Composition

63- 75 pm 46% 1.42 cc/g Sub-rounded to angular Pure Si02

The packing density o f the quartz sand was measured and the porosity calculated from this value. Flat bottomed, straight- sided Perspex cups o f a known volume were slowly filled with the quartz sand using the technique o f pluviation. The surface o f sand was leveled with the lip o f the cup with a knife blade. The sand was emptied out o f the cup on to weighing paper and

its mass was measured. Its density was simply calculated from the known volume it had occupied in the cup and its measured mass (equation 3.1).

where; V density m = mass V= volume equation 3.1

The density calculated for the porous quartz was 1.42 cc/g. A porosity o f 46% was calculated for the porous quartz from its measured density and the known density o f ‘solid’ single- crystal quartz (2.65 cc/g).

To confirm the quartz powder was pure silica a portion o f the target sand was ground to 10 microns and its chemical composition determ ined by x-ray diffraction. A plot o f intensity against 20 was produced (figure 3.3) which illustrates the sample com position was pure SiO].

Figure 3.3, Plot of 20 against intensity shows typical X-ray diffraction peaks for pure SiO], confirming the composition of the sample material.

C 3 a 24.4 81.5 0 10 20 30 40 50 60 70 80 00 100 110 120 20

A thin section o f the SiO] sand was resin mounted and exam ined under the petrological microscope. Grain shape varied from sub-rounded to angular {figures 3.4 & 3.5) and a few grain surfaces showed evidence o f m inor concoidal fracturing. Under cross-polarised light the quartz showed no indications o f strain, e.g. there was no evidence o f undulose extinction. There was no evidence o f the material having being previously shocked, and it contained no linear features that could be confused with PDFs.

n i c r c n s

Figure 3.4, Photomicrograph (cross polarized light) of sub-rounded to angular shape of control quartz grains. Photo taken at a magnification of x 100 (subsequently rescaled).

Figure 3.5, Photomicrograph (cross- polarized light) of control quartz grains with some faces showing concoidal fracturing typical of quartz. No planar or linear features are present in this pre­ shocked quartz. Photomicrograph taken at a magnification of x 400 (subsequently rescaled).

3.3.1 E xperim ental Design and A pparatus

To com pare the effect o f loading path on the extent o f shock m etam orphism , it was desirable to use container m aterials with a wide range o f shock impedances. The suitability o f the materials in terms o f ease o f machining and its properties as a sample container was a consideration and materials were limited to those with already measured shock EOS data (Kinslow, 1979; Ahrens and Johnson, 1994). To com pare shock impedances o f possible container m aterials with the shock im pedance o f the porous quartz sample, their Hugoniot curves were plotted on a P - U ^ graph (figure 3.6). From this plot it may be observed that the porous quartz with a density o f 1.43 cc/g has a com parable shock impedance to Polyethylene. Stainless steel has a much greater shock impedance than the porous quartz, and alum inium and Teflon have shock impedances interm ediate when com pared to stainless steel and porous quartz.

Figure 3.6, P - plot of Hugoniot curves for a variety of materials suitable as sample containers; tungsten (— ), stainless steel (— ), copper (— ), aluminium (— ), Teflon (— ) and polyethylene ( ) compared to the Hugoniot curves for the potential sample materials ‘non-porous’ quartz (— ) and porous quartz (— ) with a density of 1.43 cc/g. Materials with a steeper curve have a higher shock impedance than those with a shallower curve. The shock EOS used to calculate the curves for the porous quartz, quartz, Polyethylene and Tungsten are from a review of EOS measurements compiled by Ahrens and Johnson, 1994. The shock EOS for stainless steel, aluminium, copper and Teflon are from a list in Kinslow, 1970.

0 0 .5 1 1.5 2 2 .5 3 3 .5