《复合材料 Composites》课程教学资源（学习资料）第二章 增强体_glass fiber-2 The properties of glass fibres after conditioning at composite recycling temperatures

Composites: Part A 61(2014)201-208 Contents lists available at ScienceDirect campsites Composites: Part A ELSEVIER journalhomepagewww.elsevier.com/locate/compositesa The properties of glass fibres after conditioning at composite recycling Cross Mark temperatures J.L. Thomason, L Yang, R. Meier University of strathclyde, Department of Mechanical and Aerospace Engineering, 75 Montrose Street, Glasgow G1 IX). United Kingdom ARTICLE INFO A BSTRACT Results are presented on E-glass fibr 600°. Thermal 6 December 2013 to 70% strength de dane- coated fibres in revised form 28 February 2014 were relatively stable up to 250C but exhibited a precipitous drop at higher conditioning temperatures. Available online 12 march 2014 Unsized fibres exhibited a linear decrease in strength with increasing conditioning temperature. Little onditioned fibres. A simple analysis of the cumulative fibre strength probability resulted in more useful understanding than the Weibull method. The modulus of both fibre types increased linearly with onditioning temperature. Evidence was found of a slow time-dependent reduction of glass fib E Heat treat storage in an uncontrolled environment The results are discussed in terms of the changes g and bulk glass structure during heat conditioning and the role of the glass fibre water e 2014 Elsevier Ltd. All rights reserved. 1 Introduction to its original state. 7. Consequently, recycled fibres have a very poor performance to cost ratio, and in most cases are considered The disposal of end-of-life composite products in an environ- unsuitable for reprocessing and reuse as a valuable reinforcement mentally friendly manner is one of the most important challenges of composites. a breakthrough in this field could enable suck currently facing the industrial and academic composites commu- recycled glass fibres(gfS)to compete with pristine materials in nity. It is projected that by 2015 the total global production of com- many large volume composite applications. The development of posite materials will significantly exceed 10 million tons which, at an economically viable process for regenerating the properties of d- of-life, will occupy a volume of over 5 million cubic meters. thermally recycled glass fibres would have major technological Glass fibre reinforced composites account for more than 90% of societal, economical, environmental impacts. The reuse of these ll the fibre-reinforced composites currently produced. About materials could result in a huge reduction in the environment. 0% of this volume employs thermosetting matrix materials pro- impact of the glass-fibre and composites industry where the ducing composites(grP) that are difficult to recycle in an efficient replacement of pristine glass fibre products by rgF products would anner. The perspectives on this issue have been recently high- equate to a global reduction in COz production of 400,000 Tons/ lighted due to the accelerating growth in the use of such composite annum from reduced melting energy requirements alone Further materials in transportation and wind energy sectors[1-6.A num- more, such a development would also reduce the need for an ber of processes are available for recycling such composites [1, 7]. annual landfill disposal of 2 million Tons of composites. These f these possible routes, thermal recycling is probably the most developments would clearly be in line with the growing societal echnologically advanced and has been piloted in the UK and and environmental pressure to reduce the use of landfill disposal Denmark. However, nearly all options deliver recycled fibres increase the reuse of valuable raw materials resources, and reduce (which make up approximately 60% by weight of the composites) the release of Coz to the atmosphere that suffer from a lack of cost competitiveness with pristine first- Processing temperatures in the production of glass fibre are sig- pass materials. a key factor in this equation is the huge drop in nificantly higher than GRP recycling temperatures. Nevertheless, the performance of recycled glass fibre(80-90%)in comparison earlier work has indicated that the room temperature tensile strength of glass fibre can be significantly reduced by annealing at temperature as low as 150C[ 8. More recent studies have also ng author.Tel:+4401415482691;fax:+4401415525105. confirmed that room temperature glass fibre strength can be re- ss: james. thomason@strathacuk (L Thomason). ress: Technische Universitat Munchen, D-85748 Garching, German duced by exposure to temperatures in the 300-600C temperature 0.101 1359-835Xo 2014 Elsevier Ltd. All rights reserved

The properties of glass fibres after conditioning at composite recycling temperatures J.L. Thomason ⇑ , L. Yang, R. Meier 1 University of Strathclyde, Department of Mechanical and Aerospace Engineering, 75 Montrose Street, Glasgow G1 1XJ, United Kingdom article info Article history: Received 6 December 2013 Received in revised form 28 February 2014 Accepted 1 March 2014 Available online 12 March 2014 Keywords: A. Glass fibres B. Mechanical properties E. Heat treatment E. Recycling abstract Results are presented on E-glass fibre properties after thermal conditioning up to 600 C. Thermal conditioning led to up to 70% strength degradation. Tensile strength and failure strain of silane-coated fibres were relatively stable up to 250 C but exhibited a precipitous drop at higher conditioning temperatures. Unsized fibres exhibited a linear decrease in strength with increasing conditioning temperature. Little significant strength regeneration was obtained from a range of acid and silane post-treatments of heat conditioned fibres. A simple analysis of the cumulative fibre strength probability resulted in more useful understanding than the Weibull method. The modulus of both fibre types increased linearly with conditioning temperature. Evidence was found of a slow time-dependent reduction of glass fibre modulus during storage in an uncontrolled environment. The results are discussed in terms of the changes in surface coating and bulk glass structure during heat conditioning and the role of the glass fibre water content. 2014 Elsevier Ltd. All rights reserved. 1. Introduction The disposal of end-of-life composite products in an environ￾mentally friendly manner is one of the most important challenges currently facing the industrial and academic composites commu￾nity. It is projected that by 2015 the total global production of com￾posite materials will significantly exceed 10 million tons which, at end-of-life, will occupy a volume of over 5 million cubic meters. Glass fibre reinforced composites account for more than 90% of all the fibre-reinforced composites currently produced. About 60% of this volume employs thermosetting matrix materials pro￾ducing composites (GRP) that are difficult to recycle in an efficient manner. The perspectives on this issue have been recently high￾lighted due to the accelerating growth in the use of such composite materials in transportation and wind energy sectors [1–6]. A num￾ber of processes are available for recycling such composites [1,7]. Of these possible routes, thermal recycling is probably the most technologically advanced and has been piloted in the UK and Denmark. However, nearly all options deliver recycled fibres (which make up approximately 60% by weight of the composites) that suffer from a lack of cost competitiveness with pristine first￾pass materials. A key factor in this equation is the huge drop in the performance of recycled glass fibre (80–90%) in comparison to its original state [1,7]. Consequently, recycled fibres have a very poor performance to cost ratio, and in most cases are considered unsuitable for reprocessing and reuse as a valuable reinforcement of composites. A breakthrough in this field could enable such recycled glass fibres (RGFs) to compete with pristine materials in many large volume composite applications. The development of an economically viable process for regenerating the properties of thermally recycled glass fibres would have major technological, societal, economical, environmental impacts. The reuse of these materials could result in a huge reduction in the environmental impact of the glass-fibre and composites industry where the replacement of pristine glass fibre products by RGF products would equate to a global reduction in CO2 production of 400,000 Tons/ annum from reduced melting energy requirements alone. Further￾more, such a development would also reduce the need for an annual landfill disposal of 2 million Tons of composites. These developments would clearly be in line with the growing societal and environmental pressure to reduce the use of landfill disposal, increase the reuse of valuable raw materials resources, and reduce the release of CO2 to the atmosphere. Processing temperatures in the production of glass fibre are sig￾nificantly higher than GRP recycling temperatures. Nevertheless, earlier work has indicated that the room temperature tensile strength of glass fibre can be significantly reduced by annealing at temperature as low as 150 C [8]. More recent studies have also confirmed that room temperature glass fibre strength can be re￾duced by exposure to temperatures in the 300–600 C temperature http://dx.doi.org/10.1016/j.compositesa.2014.03.001 1359-835X/ 2014 Elsevier Ltd. All rights reserved. ⇑ Corresponding author. Tel.: +44 01415482691; fax: +44 01415525105. E-mail address: james.thomason@strath.ac.uk (J.L. Thomason). 1 Current address: Technische Universität München, D-85748 Garching, Germany. Composites: Part A 61 (2014) 201–208 Contents lists available at ScienceDirect Composites: Part A journal homepage: www.elsevier.com/locate/compositesa

J.L. Thomason et aL/Composites: Part A 61(2014) 201-208 range 9-12] which is typical of the many different potential grP the treatment solutions, one or two hours for acid treatment, cycling processes. Similar behaviour has also been observed in 15 min for silane solution, at room temperature and then dried silica and basalt reinforcement fibres 13, 14 We are currently en- in an oven at 110C for another 15 min gaged in research projects where the ultimate goal is the genera Single fibres were meticulously separated from the glass fibre ion of the fundamental knowledge to enable cost-effective strands avoiding fibre-fibre interactions or excessive fibre bending regeneration of the mechanical properties of glass fibres produced as much as possible. Individual fibres were glued onto a card tabs from thermal recycling of glass reinforced thermoset composites with a central window cut out to matched the desired gauge length such as wind turbine blades. In this paper we report on the influ- for the test Card frames were cut from 250 g/m grade card and ence of thermal conditioning, at temperatures typical for GRP recy- single fibres were fixed to the card at both sides of the window ling up to 600C, on the properties of water sized and silane sized using Loctite TM Gel Superglue. A Nikon Epiphot Inverted optical E-glass fibres. We also report initial results of a study on the use of microscope was used at 200x magnification to obtain a digital acid and silane treatments on the strength of heat-treated fibres. photo of each fibre. The cross-sectional area was calculated from individual average fibre diameters measured at five points along gauge length During handling of the fibre in the microscope, care was taken to avoid fibre damage through contact with the microscope objective. Single fibre tensile properties we Boron free E-glass fibres supplied by Owens Corning-Vetrotex mined following ASTM C1557-03 using an Instron 3342 vere investigated in this work. The fibre rovings were produced testing machine equipped with a 10N load cell. Sampl on a pilot scale bushing and were received as 20 kg continuous sin- length was 20 mm for both fibre types and approximately 75 fibres gle end square edge packages. The rovings had a nominal tex of were tested at each condition. The tensile testing strain rate used 1200 g/m and a nominal fibre diameter of 17 um. No sizing was ap- was 1.5% min and all the tests were carried out at room tempera plied to the water finished fibres which had only been sprayed ture and 50% relative humidity. Only the tests where the sample using the normal water prepad cooling sprays under the bushing, broke along the gauge length at a distance greater than 3 mm from these samples are referred to as water sized or unsized. The APs the clamps were used for further data processing coated fibres were coated with a normal rotating cylinder sizing plicator containing a 1% volume r-aminopropy APS) hydrolysed solution in deionized water. All fibre packages 3 Results and discussion were subsequently dried at 105C for 24 h. The fibres were used as received from the manufacturer. heat treatment of both fibre 3.1. fibre diameter distribution types was conducted simultaneously to obtain samples with iden- ical thermal history. 300 mm lengths of silane sized and water Over the course of the investigation the diameters of more than sized fibre strand with no visible damage were removed from the 1200 individual glass fibres were measured using optical micros- inside of the roving packages. The glass fibre strands were sus- copy. The results are summarised in Table 1 which reveals the rel- pended on a specially constructed jig preventing any contact with, atively large distribution fibre diameters present in and therefore damage to, the fibres(see Fig. 1). Heat conditioning commercially produced glass fibre reinforcements. Fibre diameter ras carried out in a Carbolite LHT6 high temperature oven for is an important parameter in defining the final performance of fi- 15 min at temperatures in the range 250-600C The jig was then bre reinforced composites and is related to several parameters removed from the oven and the samples allowed to cool naturally including the internal diameter of the bushing tip, the velocity of to room temperature. molten glass flow, and attenuation rate 15. For a given glass-fibre For chemical post-treatment ACS reagent 37% hydrochloric acid manufacturing configuration it is the variation of the molten gla (HCI). APS. y-glycidyloxypropyltrimethoxysilane(GPS)and merca- flow that mainly accounts for fibre diameter distribution. The va ptopropyltriethoxysilane(sPs)were supplied by Sigma Aldrich in iation in the viscosity of glass melt is believed to be caused by the the UK. 10 v% HCl was prepared by diluting concentrated HCl with temperature distribution across the bushing 15 In the deionised water. 1% volume silane solution was prepared by mix- gle fibre strength and modulus these fibre diameters are used to ing the silane with deionised water. APS was hydrolysed without calculate the fibre cross section. Examining the values for mini- PH adjustment, all other silanes were hydrolysed in water whose mum and maximum fibre diameter in Table 1 it can be seen that pH value had been adjusted to 5. 1-5.3 prior to mixing One silane using only a manufacturers nominal fibre diameter or a single blend was created by mixing the APS and SPs solutions in equal average value of the distribution can lead to up to a 78%error mounts. The solution were left for 24 h to ensure full hydrolysis. the value of fibre cross section. This would translate directly into Heat treated glass fibre bundles were completely immersed in errors of similar magnitude in the fibre modulus and strength. This Fig. 1. Illustration of fibre suspension rig and rig positioned in furnace. For interpretation of the references to colour in this figure legend, the reader is referred to the web ersion of th

range [9–12] which is typical of the many different potential GRP recycling processes. Similar behaviour has also been observed in silica and basalt reinforcement fibres [13,14]. We are currently en￾gaged in research projects where the ultimate goal is the genera￾tion of the fundamental knowledge to enable cost-effective regeneration of the mechanical properties of glass fibres produced from thermal recycling of glass reinforced thermoset composites such as wind turbine blades. In this paper we report on the influ￾ence of thermal conditioning, at temperatures typical for GRP recy￾cling up to 600 C, on the properties of water sized and silane sized E-glass fibres. We also report initial results of a study on the use of acid and silane treatments on the strength of heat-treated fibres. 2. Experimental Boron free E-glass fibres supplied by Owens Corning-Vetrotex were investigated in this work. The fibre rovings were produced on a pilot scale bushing and were received as 20 kg continuous sin￾gle end square edge packages. The rovings had a nominal tex of 1200 g/m and a nominal fibre diameter of 17 lm. No sizing was ap￾plied to the water finished fibres which had only been sprayed using the normal water prepad cooling sprays under the bushing, these samples are referred to as water sized or unsized. The APS coated fibres were coated with a normal rotating cylinder sizing applicator containing a 1% volume c-aminopropyltriethoxysilane (APS) hydrolysed solution in deionized water. All fibre packages were subsequently dried at 105 C for 24 h. The fibres were used as received from the manufacturer. Heat treatment of both fibre types was conducted simultaneously to obtain samples with iden￾tical thermal history. 300 mm lengths of silane sized and water sized fibre strand with no visible damage were removed from the inside of the roving packages. The glass fibre strands were sus￾pended on a specially constructed jig preventing any contact with, and therefore damage to, the fibres (see Fig. 1). Heat conditioning was carried out in a Carbolite LHT6 high temperature oven for 15 min at temperatures in the range 250–600 C. The jig was then removed from the oven and the samples allowed to cool naturally to room temperature. For chemical post-treatment ACS reagent 37% hydrochloric acid (HCl), APS, c-glycidyloxypropyltrimethoxysilane (GPS) and merca￾ptopropyltriethoxysilane (SPS) were supplied by Sigma Aldrich in the UK. 10 v% HCl was prepared by diluting concentrated HCl with deionised water. 1% volume silane solution was prepared by mix￾ing the silane with deionised water. APS was hydrolysed without pH adjustment, all other silanes were hydrolysed in water whose pH value had been adjusted to 5.1–5.3 prior to mixing. One silane blend was created by mixing the APS and SPS solutions in equal amounts. The solution were left for 24 h to ensure full hydrolysis. Heat treated glass fibre bundles were completely immersed in the treatment solutions, one or two hours for acid treatment, 15 min for silane solution, at room temperature and then dried in an oven at 110 C for another 15 min. Single fibres were meticulously separated from the glass fibre strands avoiding fibre–fibre interactions or excessive fibre bending as much as possible. Individual fibres were glued onto a card tabs with a central window cut out to matched the desired gauge length for the test. Card frames were cut from 250 g/m2 grade card and single fibres were fixed to the card at both sides of the window using Loctite™ Gel Superglue. A Nikon Epiphot Inverted optical microscope was used at 200 magnification to obtain a digital photo of each fibre. The cross-sectional area was calculated from individual average fibre diameters measured at five points along the gauge length. During handling of the fibre in the microscope, care was taken to avoid fibre damage through contact with the microscope objective. Single fibre tensile properties were deter￾mined following ASTM C1557-03 using an Instron 3342 universal testing machine equipped with a 10 N load cell. Sample gauge length was 20 mm for both fibre types and approximately 75 fibres were tested at each condition. The tensile testing strain rate used was 1.5%/min and all the tests were carried out at room tempera￾ture and 50% relative humidity. Only the tests where the sample broke along the gauge length at a distance greater than 3 mm from the clamps were used for further data processing. 3. Results and discussion 3.1. Fibre diameter distribution Over the course of the investigation the diameters of more than 1200 individual glass fibres were measured using optical micros￾copy. The results are summarised in Table 1 which reveals the rel￾atively large distribution in fibre diameters present in commercially produced glass fibre reinforcements. Fibre diameter is an important parameter in defining the final performance of fi- bre reinforced composites and is related to several parameters including the internal diameter of the bushing tip, the velocity of molten glass flow, and attenuation rate [15]. For a given glass-fibre manufacturing configuration it is the variation of the molten glass flow that mainly accounts for fibre diameter distribution. The var￾iation in the viscosity of glass melt is believed to be caused by the temperature distribution across the bushing [15]. In the case of sin￾gle fibre strength and modulus these fibre diameters are used to calculate the fibre cross section. Examining the values for mini￾mum and maximum fibre diameter in Table 1 it can be seen that using only a manufacturers nominal fibre diameter or a single average value of the distribution can lead to up to a 78% error in the value of fibre cross section. This would translate directly into errors of similar magnitude in the fibre modulus and strength. This Fig. 1. Illustration of fibre suspension rig and rig positioned in furnace. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 202 J.L. Thomason et al. / Composites: Part A 61 (2014) 201–208

J. L. Thomason et aL/Composites: Part A 61 (2014) 201-208 Fibre diameter analysis. Unsized 1761 Standard deviation (um) Minimum diameter(um) 15.05 Maximum diameter (um 23.06 9巴 1.5 n the measured diameters of glass fibres emphasise termine individual fibre diameters prior to each sin- t and not use average or datasheet values Treatment Temperature('C) 3. 2. Single fibre testing results Fig 3 Influence of heat treatment temperature on room temperature glass fibre average failure strain (a water sized, o APS sized) The results for the average single fibre strength of silane sized and water sized fibres after heat treatment are shown in Fig. 2 strength reduction after elevated temperature conditioning it and the fibre strain at failure is shown in Fig 3. The results indicate seems probable that there also exists a strength reduction mecha that thermal conditioning caused a considerable strength reduc nism related to a fundamental change in the glass itself. This con- tion for both silane-sized and water-sized fibre samples, with a loss clusion is supported by the results of a deeper investigation of the of over 70% of the original strength in the case of conditioning at relationship between the strength of thermally conditioned glas 600C. It can be seen that both glass fibre types reduce in strength, fibres and the degradation of the silane coating which has recently with the silane sized glass falling by a greater percentage of its ori- been reported (181 ginal strength. The water sized fibres exhibit a fairly linear reduc It is well known that the commonly used method to determine tion in room temperature strength with increasing conditioning single fibre modulus delivers results that are directly dependent on temperature. The strength of silane-sized glass appears relatively the gauge length tested. The modulus dependence on gauge length stable at low temperatures but exhibits a threshold (at appre is a phenomenon related to the use of the testing machine cross- mately 250C) above which a precipitous reduction in fibre head position to approximate the fibre strain that does not take strength and strain to failure occurs. These results appear to agree into account the contribution of the strain of the other components well with results from other investigators [8-12 ].Consequently it in the testing system. All modulus values reported here have been appears that the average strength of glass fibres is strongly influ- corrected for the test system compliance using a value of enced by thermal conditioning at temperatures and times which .0346 mm/GPa, obtained from Fig 4 using data from a recently re- nay commonly be experienced during the thermal processing of ported gauge length study of the same fibres [16]. The results for end-of-life composite materials in ord er to obtained red the average room temperature fibre modulus (error bars show 7]. These effects appear to be occurring at temperatures signifi- 95% confidence limits)of unsized and silane sized glass fibres are cantly lower than the 1150-1250C temperature range for the shown as a function of the sample conditioning temperature in manufacture of glass fibre and also well below 759C the glass Fig. 5. It can be seen that both samples show a significant trend transition temperature of boron free E-glass 19 for room temperature modulus being increased by a short elevated Further important conclusions that can be drawn from Figs. 2 temperature conditioning. It is also clear in Fig. 5 that both samples and 3 include the clear role of sizing in protecting the strength let show an approximately linear increase in room quent loss of protection, during thermal conditioning, of the Furthermore, although both samples started with approximately organic part of the polysiloxane layer on the fibre surface, very the same level of tensile modulus. the modulus of the water ikely contributes to the loss of average fibre strength in the silane fibres was found to be significantly higher(at 95% confidence level) sized fibres However since the water sized fibres also exhibit a y=00130x+0.0343 y=0.0127x+0.0349 Treatment Temperature('C) ge Length(mm) of heat treatment temperature on perature glass fibre Fig, 4 Gauge length dependence of fibre modulus for compliance correction factor (▲ water sized.● APS sized) (▲ water sized,· APS sized)

large range in the measured diameters of glass fibres emphasises the need to determine individual fibre diameters prior to each sin￾gle fibre test and not use average or datasheet values. 3.2. Single fibre testing results The results for the average single fibre strength of silane sized and water sized fibres after heat treatment are shown in Fig. 2 and the fibre strain at failure is shown in Fig. 3. The results indicate that thermal conditioning caused a considerable strength reduc￾tion for both silane-sized and water-sized fibre samples, with a loss of over 70% of the original strength in the case of conditioning at 600 C. It can be seen that both glass fibre types reduce in strength, with the silane sized glass falling by a greater percentage of its ori￾ginal strength. The water sized fibres exhibit a fairly linear reduc￾tion in room temperature strength with increasing conditioning temperature. The strength of silane-sized glass appears relatively stable at low temperatures but exhibits a threshold (at approxi￾mately 250 C) above which a precipitous reduction in fibre strength and strain to failure occurs. These results appear to agree well with results from other investigators [8–12]. Consequently it appears that the average strength of glass fibres is strongly influ￾enced by thermal conditioning at temperatures and times which may commonly be experienced during the thermal processing of end-of-life composite materials in order to obtained recycled fibres [7]. These effects appear to be occurring at temperatures signifi- cantly lower than the 1150–1250 C temperature range for the manufacture of glass fibre and also well below 759 C the glass transition temperature of boron free E-glass [19]. Further important conclusions that can be drawn from Figs. 2 and 3 include the clear role of sizing in protecting the strength lev￾els in glass fibres. It seems likely that the degradation and conse￾quent loss of protection, during thermal conditioning, of the organic part of the polysiloxane layer on the fibre surface, very likely contributes to the loss of average fibre strength in the silane sized fibres. However, since the water sized fibres also exhibit a strength reduction after elevated temperature conditioning it seems probable that there also exists a strength reduction mecha￾nism related to a fundamental change in the glass itself. This con￾clusion is supported by the results of a deeper investigation of the relationship between the strength of thermally conditioned glass fibres and the degradation of the silane coating which has recently been reported [18]. It is well known that the commonly used method to determine single fibre modulus delivers results that are directly dependent on the gauge length tested. The modulus dependence on gauge length is a phenomenon related to the use of the testing machine cross￾head position to approximate the fibre strain that does not take into account the contribution of the strain of the other components in the testing system. All modulus values reported here have been corrected for the test system compliance using a value of 0.0346 mm/GPa, obtained from Fig. 4 using data from a recently re￾ported gauge length study of the same fibres [16]. The results for the average room temperature fibre modulus (error bars show 95% confidence limits) of unsized and silane sized glass fibres are shown as a function of the sample conditioning temperature in Fig. 5. It can be seen that both samples show a significant trend for room temperature modulus being increased by a short elevated temperature conditioning. It is also clear in Fig. 5 that both samples show an approximately linear increase in room temperature mod￾ulus with conditioning temperature above approximately 200 C. Furthermore, although both samples started with approximately the same level of tensile modulus, the modulus of the water sized fibres was found to be significantly higher (at 95% confidence level) Table 1 Fibre diameter analysis. Unsized APS sized Average diameter (lm) 18.18 17.61 Standard deviation (lm) 1.28 1.53 Number of fibres 537 719 Confidence level (95.0%) 0.11 0.11 Minimum diameter (lm) 15.05 13.88 Maximum diameter (lm) 23.06 23.52 0.5 1.0 1.5 2.0 2.5 0 100 200 300 400 500 600 Single Fibre Strength (GPa) Treatment Temperature (°C) Fig. 2. Influence of heat treatment temperature on room temperature glass fibre average strength (N water sized, d APS sized). 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 100 200 300 400 500 600 Single Fibre Failure Strain (%) Treatment Temperature (°C) Fig. 3. Influence of heat treatment temperature on room temperature glass fibre average failure strain (N water sized, d APS sized). y = 0.0130x + 0.0343 y = 0.0127x + 0.0349 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 20 40 60 80 Length/Modulus (mm/GPa) Gauge Length (mm) Fig. 4. Gauge length dependence of fibre modulus for compliance correction factor (N water sized, d APS sized). J.L. Thomason et al. / Composites: Part A 61 (2014) 201–208 203

J.L. Thomason et aL/Composites: Part A 61(2014) 201-208 Measured 20 months earlier(Ref xx) Silane sized23° 三5 300400 Treatment Temperature('C) 0.5 average modulus(▲ water sized,● APS sized) than that of the silane coated fibres after elevated temperature conditioning. Hence it appears that the silane coating impedes sis of the data from all fibres and conditions as all sample strength the mechanism that leads to an increase in modulus of the fibres data produced Weibull plots similar in form to one of these two during the hot conditioning. Results for the unconditioned fibre examples. At higher strength values the data for the 600C condi modulus for both samples from a previous investigation measured tioned fibres mainly follow a linear relationship typical of a uni- approximately 20 months prior to the current measurements are modal flaw distribution. However, at lower strengths the data also presented in Fig. 5. It is noted that the modulus values for both exhibit a downward deviation in curvature that is a symptom of fibre samples are significantly lower in this investigation. the experimental inability to access very low strength values as discussed above. Consequently the data are poorly fitted by a 3.3. Discussion of fibre strength results straight line (indicative of the simplest two parameter Weibull dis- tribution but can be well fitted, as shown, by a more general three Data from single fibre strength measurement are commonly parameter Weibull distribution. Nevertheless, it has been shown subjected to further detailed analysis using Weibull methods that it is actually more accurate to fit the data with a two-param- 20. These details of the Weibull method are well documented eter model where the calculated values below the lower experi and will not be reproduced here[17, 20. In some cases, experimen- mental limit L have been removed [17 The data for the APs tal single fibre tensile strength data can be well described using a sized fibres without any heat conditioning exhibited a more com wo parameter unimodal Weibull analysis. However, curvature in plex fibre strength distribution and appear to exhibit three distinct the Weibull plot of experimental single fibre data is frequently ob- regions in Fig. 6. This could be interpreted as indicating that the served at low strength values [ 17]. A recent study has demon- failure of silane coated glass fibres is controlled by two, or more strated that this deviation can be well explained by the existence distinct types of flaw population. The overall failure distribution of an experimental lower strength limit(or)below which the fibre is then the result of the interaction of these flaw populations. is not strong enough to survive the pretest sample preparation and which may be described by the multimodal Weibull distribution [17. Consequently, the lowest strength values of the esti- and precludes the use of the simple linear fitting to obtain the wei- Weibull distribution do not appear in the experimental bull parameters. It is possible to fit the results for the APS sized fi- data In such cases it is necessary to revert to the original three bres using a partially concurrent bimodal Weibull distribution 12 parameter form of Weibull distribution 17. Unfortunately this where the lower strength part of the distribution also includes a phenomenon seriously complicates attempts to use Weibull anal- non-zero value of oT. However, it bears commenting that this re- ysis of single fibre strength data to distil useful information about quires the reliable estimation of six adjustable parameters. the materials science of the tested single fibres. The experimentally Although further discussion and analysis of all fibre strength data reflected lower strength limit is operator sensitive, sample prepa- is certainly possible, one has to question whether all of the exper Ition dependent, and test procedure dependent as well as being imental and theoretical uncertainties discussed above warrant any related to the material properties. Consequently, although average attempts at interpretation of this type of Weibull analysis in terms values for or can be obtained the value of or is not a constant of real material parameters but can be different for each individual fibre test this A more robust analysis method that requires a less complex situation can become even more complex when further structure theoretical background is simply to plot the cumulative probability is observed in a Weibull plot, often interpreted as evidence for of fibre failure(Pf)as a function of fibre strength as shown in Figs multiple flaw populations in the fibres under test [12, 16, 17. This and 8 for the two fibre types in this investigation. Fig. 7 shows the is frequently observed in the distribution of strength from sized data for the water sized fibres. It can be observed that there is con- and carbon fibres and analysis using different forms of bimo- siderable separation between the unsized fibres conditioned above dal Weibull strength distribution has been attempted. Thomason 300C and the as-received fibres. The region between these two et al. recently presented data on the aps sized fibres used in this groups appears to be spanned by the distribution for the fibres con- investigation indicating that the partially concurrent bimodal Wei- ditioned at 250C. It can also be observed that all the heat condi bull distribution gave the most realistic model for these results tioned fibres have a similar low strength distribution for the first 12 25-30% of the distribution. One possible explanation for these Typical examples are presented in Fig. 6 which shows Weibull observations is that the low strength region of fibre strength of plots of the strength data for the APS fibres before heat condition- the heat conditioned fibres is a result of further mechanical dam g and for the water-sized fibres after heat conditioning at 600C. age to the unsized fibres caused by the handling of the fibres in This covers the full range of behaviour observed in Weibull analy- the process of heat conditioning. The strength of any stronger

than that of the silane coated fibres after elevated temperature conditioning. Hence it appears that the silane coating impedes the mechanism that leads to an increase in modulus of the fibres during the hot conditioning. Results for the unconditioned fibre modulus for both samples from a previous investigation measured approximately 20 months prior to the current measurements are also presented in Fig. 5. It is noted that the modulus values for both fibre samples are significantly lower in this investigation. 3.3. Discussion of fibre strength results Data from single fibre strength measurement are commonly subjected to further detailed analysis using Weibull methods [20]. These details of the Weibull method are well documented and will not be reproduced here [17,20]. In some cases, experimen￾tal single fibre tensile strength data can be well described using a two parameter unimodal Weibull analysis. However, curvature in the Weibull plot of experimental single fibre data is frequently ob￾served at low strength values [17]. A recent study has demon￾strated that this deviation can be well explained by the existence of an experimental lower strength limit (rT) below which the fibre is not strong enough to survive the pretest sample preparation and handling [17]. Consequently, the lowest strength values of the esti￾mated Weibull distribution do not appear in the experimental data. In such cases it is necessary to revert to the original three parameter form of Weibull distribution [17]. Unfortunately this phenomenon seriously complicates attempts to use Weibull anal￾ysis of single fibre strength data to distil useful information about the materials science of the tested single fibres. The experimentally reflected lower strength limit is operator sensitive, sample prepa￾ration dependent, and test procedure dependent as well as being related to the material properties. Consequently, although average values for rT can be obtained, the value of rT is not a constant parameter but can be different for each individual fibre test. This situation can become even more complex when further structure is observed in a Weibull plot, often interpreted as evidence for multiple flaw populations in the fibres under test [12,16,17]. This is frequently observed in the distribution of strength from sized glass and carbon fibres and analysis using different forms of bimo￾dal Weibull strength distribution has been attempted. Thomason et al. recently presented data on the APS sized fibres used in this investigation indicating that the partially concurrent bimodal Wei￾bull distribution gave the most realistic model for these results [12]. Typical examples are presented in Fig. 6 which shows Weibull plots of the strength data for the APS fibres before heat condition￾ing and for the water-sized fibres after heat conditioning at 600 C. This covers the full range of behaviour observed in Weibull analy￾sis of the data from all fibres and conditions as all sample strength data produced Weibull plots similar in form to one of these two examples. At higher strength values the data for the 600 C condi￾tioned fibres mainly follow a linear relationship typical of a uni￾modal flaw distribution. However, at lower strengths the data exhibit a downward deviation in curvature that is a symptom of the experimental inability to access very low strength values as discussed above. Consequently the data are poorly fitted by a straight line (indicative of the simplest two parameter Weibull dis￾tribution) but can be well fitted, as shown, by a more general three parameter Weibull distribution. Nevertheless, it has been shown that it is actually more accurate to fit the data with a two-param￾eter model where the calculated values below the lower experi￾mental limit rL have been removed [17]. The data for the APS sized fibres without any heat conditioning exhibited a more com￾plex fibre strength distribution and appear to exhibit three distinct regions in Fig. 6. This could be interpreted as indicating that the failure of silane coated glass fibres is controlled by two, or more, distinct types of flaw population. The overall failure distribution is then the result of the interaction of these flaw populations, which may be described by the multimodal Weibull distribution and precludes the use of the simple linear fitting to obtain the Wei￾bull parameters. It is possible to fit the results for the APS sized fi- bres using a partially concurrent bimodal Weibull distribution [12] where the lower strength part of the distribution also includes a non-zero value of rT. However, it bears commenting that this re￾quires the reliable estimation of six adjustable parameters. Although further discussion and analysis of all fibre strength data is certainly possible, one has to question whether all of the exper￾imental and theoretical uncertainties discussed above warrant any attempts at interpretation of this type of Weibull analysis in terms of real material parameters. A more robust analysis method that requires a less complex theoretical background is simply to plot the cumulative probability of fibre failure (PF) as a function of fibre strength as shown in Figs. 7 and 8 for the two fibre types in this investigation. Fig. 7 shows the data for the water sized fibres. It can be observed that there is con￾siderable separation between the unsized fibres conditioned above 300 C and the as-received fibres. The region between these two groups appears to be spanned by the distribution for the fibres con￾ditioned at 250 C. It can also be observed that all the heat condi￾tioned fibres have a similar low strength distribution for the first 25–30% of the distribution. One possible explanation for these observations is that the low strength region of fibre strength of the heat conditioned fibres is a result of further mechanical dam￾age to the unsized fibres caused by the handling of the fibres in the process of heat conditioning. The strength of any stronger 68 70 72 74 76 78 80 0 100 200 300 400 500 600 Fibre Modulus (GPa) Treatment Temperature (°C) Measured 20 months earlier (Ref xx) Fig. 5. Influence of heat treatment temperature on room temperature glass fibre average modulus (N water sized, d APS sized). -6 -5 -4 -3 -2 -1 0 1 2 3 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 ln (-ln(1-P)) ln(σ) [GPa] Silane Sized 23 oC Water Sized 600 oC Fig. 6. Weibull plot of single fibre strength (+ water sized fibres after 600 C conditioning, unconditioned APS sized fibres. 204 J.L. Thomason et al. / Composites: Part A 61 (2014) 201–208

J. L. Thomason et aL/Composites: Part A 61 (2014) 201-208 The effect of chemical treatments on the strength of 450"C conditioned glass fibre. ment (h) type strength(GPa) confidence 06 04 0.0020.40.60.81.012141.61.8 APS: SPS 0.81 004 Fibre Strength(GPa) Fig. 7. Cumulative failure probability for water sized single fibres after thermal nditioning(◆23°口250°0300°x380°+450 00°C) iture dependent structural effects observed with the water sized fibre data will also increasingly affect the results for the APS sized fibres conditioned at increasingly higher temperatures 3.4. Infiuence of chemical post-treatment on the glass fibre strength The effects of various chemical post-treatments on the strength of 450C heat-conditioned glass fibres are summarised in the Table 2. Direct silanisation with either a single silane(Aps or GPS) or the APS: SPs silane blend did not result in any significant increase in the average fibre strength of heat-treated fibres. the expectation for restoring fibre strength through simple silanisation originated from the critical role that silane coupling agents play in the manufacture of glass fibre reinforcements. It is well known that these silanes have the ability to help maintain fibre strength[16.It has been recently demonstrated that strength retention caused by Fibre Strength(GPa) silane deposition on glass fibres during manufacture can be ex- plained from surface protection viewpoint [16. However, it has Fig 8: Cumulative failure probability for Aps sized single fibres after thermal also been suggested by some authors that silanes can'healsurface flaws by patching the crack tips, mitigating stress concentration water-sized fibres is then also reduced in greater degree w during loading 21, 22 It is reasonable to suggest that polysiloxane bonds with high cross-link density need to be formed between increasing heat conditioning temperature possibly indicating the deposited silane molecules and the glass surface if any significant presence of some temperature dependent structural change in effect is to be achieved through such a healing mechanism. one these fibres 12, 18, 19). of possibilities for the failure of direct silanisation in strength In the case of the APs sized fibres the data in Fig. 8 indicate regeneration of the heat-treated glass fibre may be the deactiva somewhat different trends in PE. The as received fibres show a tion of glass surface by exposure to elevated temperatures. It has broad strength distribution that appears to be split approximately been shown with both amorphous silica and glass fibre that severe 50: 50 in low and high strength regions. Initially heat conditioning dehydration can occur when subjected to thermal treatment above appears to affect the high strength part of this distribution, shifting 400C in air [23, 24]. While it has been suggested that complete it to lower values with increasing heat conditioning from 250C to rehydration of such calcinated surfaces may be very difficult.it 980C. The lowest 30% of these three distributions appear to be has been reported that acid treatment with HCl can successfully re- unaffected. Heat conditioning at 450C and above, appears to shift store hydroxyl groups on the surface of glass fibre heat-treated at all fibres to lower strengths and removes t the appearance of bimo- 450C [24, 25]. In the present work, we also attempted this acid dality. Building on the discussion of the water sized fibre results reactivation of the fibre surface prior to a further attempt at silani- this could be interpreted as indicating that the APS coating protects sation. The results in Table 2 indicate that HCl treatment itself does the fibres from the possibility of mechanical damage during the not significantly change the strength of heat-treated fibres and a handling of the fibres in the process of heat conditioning. Conse- similar lack of significant changes was found for samples silanised quently there is no reduction in strength of the fibres in the low after the acid reactivation treatment, as shown b les g-K. trength region of the distribution. Nevertheless, heat conditioning Several possible explanations exist which may account for this lack will degrade the organic part of the silane coating [10, 12, 18]. It of strength improvement. Although it has been reported that E- seems likely that the high strength fraction of the fibres are ini- glass fibre calcinated at 450 C can be reactivated with a significant tially well protected by the silane and so it is the high strength increase of surface silanol groups using acidic conditions, the num- raction of the distribution that is initially affected by the thermal ber of silanol groups may be still insufficient to form a tight net conditioning at lower(250-380C) temperatures. At 450C and work in the vicinity of the critical flaws. In fact, one may we above it has been shown that the organic part of the silane is fully question whether such a strong flaw repair patch could ever be degraded resulting in a further reduction in the strength of the formed with relatively large silane molecules. Other possibilitie high strength fibres [18]. It can also be supposed that the temper- may be associated with compositional change in the glass fibre

water-sized fibres is then also reduced in greater degree with increasing heat conditioning temperature possibly indicating the presence of some temperature dependent structural change in these fibres [12,18,19]. In the case of the APS sized fibres the data in Fig. 8 indicate somewhat different trends in PF. The as received fibres show a broad strength distribution that appears to be split approximately 50:50 in low and high strength regions. Initially heat conditioning appears to affect the high strength part of this distribution, shifting it to lower values with increasing heat conditioning from 250 C to 380 C. The lowest 30% of these three distributions appear to be unaffected. Heat conditioning at 450 C and above, appears to shift all fibres to lower strengths and removes the appearance of bimo￾dality. Building on the discussion of the water sized fibre results this could be interpreted as indicating that the APS coating protects the fibres from the possibility of mechanical damage during the handling of the fibres in the process of heat conditioning. Conse￾quently there is no reduction in strength of the fibres in the low strength region of the distribution. Nevertheless, heat conditioning will degrade the organic part of the silane coating [10,12,18]. It seems likely that the high strength fraction of the fibres are ini￾tially well protected by the silane and so it is the high strength fraction of the distribution that is initially affected by the thermal conditioning at lower (250–380 C) temperatures. At 450 C and above it has been shown that the organic part of the silane is fully degraded resulting in a further reduction in the strength of the high strength fibres [18]. It can also be supposed that the temper￾ature dependent structural effects observed with the water sized fibre data will also increasingly affect the results for the APS sized fibres conditioned at increasingly higher temperatures. 3.4. Influence of chemical post-treatment on the glass fibre strength The effects of various chemical post-treatments on the strength of 450 C heat-conditioned glass fibres are summarised in the Table 2. Direct silanisation with either a single silane (APS or GPS) or the APS:SPS silane blend did not result in any significant increase in the average fibre strength of heat-treated fibres. The expectation for restoring fibre strength through simple silanisation originated from the critical role that silane coupling agents play in the manufacture of glass fibre reinforcements. It is well known that these silanes have the ability to help maintain fibre strength [16]. It has been recently demonstrated that strength retention caused by silane deposition on glass fibres during manufacture can be ex￾plained from surface protection viewpoint [16]. However, it has also been suggested by some authors that silanes can ‘heal’ surface flaws by patching the crack tips, mitigating stress concentration during loading [21,22]. It is reasonable to suggest that polysiloxane bonds with high cross-link density need to be formed between deposited silane molecules and the glass surface if any significant effect is to be achieved through such a healing mechanism. One of possibilities for the failure of direct silanisation in strength regeneration of the heat-treated glass fibre may be the deactiva￾tion of glass surface by exposure to elevated temperatures. It has been shown with both amorphous silica and glass fibre that severe dehydration can occur when subjected to thermal treatment above 400 C in air [23,24]. While it has been suggested that complete rehydration of such calcinated surfaces may be very difficult, it has been reported that acid treatment with HCl can successfully re￾store hydroxyl groups on the surface of glass fibre heat-treated at 450 C [24,25]. In the present work, we also attempted this acid reactivation of the fibre surface prior to a further attempt at silani￾sation. The results in Table 2 indicate that HCl treatment itself does not significantly change the strength of heat-treated fibres and a similar lack of significant changes was found for samples silanised after the acid reactivation treatment, as shown by samples G–K. Several possible explanations exist which may account for this lack of strength improvement. Although it has been reported that E￾glass fibre calcinated at 450 C can be reactivated with a significant increase of surface silanol groups using acidic conditions, the num￾ber of silanol groups may be still insufficient to form a tight net￾work in the vicinity of the critical flaws. In fact, one may well question whether such a strong flaw repair patch could ever be formed with relatively large silane molecules. Other possibilities may be associated with compositional change in the glass fibre 0.0 0.2 0.4 0.6 0.8 1.0 Cumulative Probability Fibre Strength (GPa) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Fig. 7. Cumulative failure probability for water sized single fibres after thermal conditioning ( 23 C, h 250 C, s 300 C, 380 C, + 450 C, N 600 C). 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Cumulative Probability Fibre Strength (GPa) Fig. 8. Cumulative failure probability for APS sized single fibres after thermal conditioning ( 23 C, h 250 C, s 300 C, 380 C, + 450 C, N 600 C). Table 2 The effect of chemical treatments on the strength of 450 C conditioned glass fibre. Sample Duration of acid treatment (h) Silane type Ave. tensile strength (GPa) 95% Confidence limit A 0 – 0.68 0.05 B 0 APS 0.86 0.07 C 0 GPS 0.73 0.08 D 0 APS:SPS 0.64 0.05 E 1 – 0.58 0.04 F 2 – 0.74 0.04 G 1 APS 0.75 0.05 H 2 APS 0.74 0.04 I 1 GPS 0.8 0.06 J 2 GPS 0.79 0.05 K 1 APS:SPS 0.81 0.04 J.L. Thomason et al. / Composites: Part A 61 (2014) 201–208 205

J.L. Thomason et aL/Composites: Part A 61(2014)201-208 Table 3 The effect of chemical treatments on the strength of unsized glass fibres. 3 15+Water sized -e H-Glass(ref 11) limit 110 000111 APS: SPS 1.38 105 100}· surface. Both thermal conditioning and acid treatment have the po- tential to significantly change surface composition, which may then change the reactivity with the silanes in comparison to the Treatment Temperature(c) surface of pristine glass fibre. To further investigate these possibil- Fig 9. Influence of heat treatment temperature on room temperature glass fibre ities, similar chemical treatments were applied to the unsized glass relative modulu fibres without prior heat conditioning. The results for these chem- ically treated fibres listed in Table 3 also indicate only minor ef temperature conditioning and recorded a much larger relative in- fects of acid and or silane treatment on the strength of unsized crease in modulus at higher conditioning temperatures In the case fibres that had no received thermal conditioning whatsoever. The of our results it appears that the uncoated fibre results agree well results presented in Table 2 and 3 raise further doubts on the valid- with those of Otto whereas the results for the APS coated fibres ap- ity of the flaw healing theory but more importantly indicate that pear to match those of Korwin-Edson et al. with an initial drop in further work is needed to improve understanding of the role of modulus. In a recent paper, Yang and Thomason discussed the no- post-silanisation in the strength of glass fibres. vel use of a thermo-mechanical analyser to simultaneously mea- sure the thermal expansion coefficient, densification, modulus of single glass fibres(the same APS coated fibres used in this investi 3.5. Discussion of fibre modulus results gation)and also to thermally condition and measure room temper ature modulus in the same instrument [19]. Their results for room Although the reduction of room temperature glass fibre temperature modulus as a function of conditioning temperature rength due to an elevated temperature conditioning has been ob-(for 30 min)are also compared in Fig 9. These data appear to show less discussion of the effects on the fibre modulus. Feih et al. ob- low conditioning temperatures and then a rapid increase in modu- served similar strength losses in heat conditioned E-glass fibres lus with conditioning above 300C. They suggested that these as reported here. However they reported no significant change in observations could be explained by a combination of structural the modulus of their fibres 10. Kennerley et al. also reported no relaxation in the glass fibre combined with desorption of water apparent change in the modulus of E-glass fibres recycled out of from within the fibre [19 Water can be present in significant RP at temperatures from 450 to 600C [26]. However, Otto at quantities in glass and can act as plasticiser. Removal of the water wens Corning reported that the modulus of E-glass fibre [27 at from the network may stiffen the structure and contribute to the room temperature increased systematically from 75 GPa up to increase of the fibre modulus. 6 GPa after conditioning for four hours at a temperature of Glasses manufactured using standard melt technologies always 600C. This was related to changes in the density of the glass contain residual water. [28 Water can exist in, and diffuse due to densification during conditioning. Glass fibre density is low- through, glass as both molecular water(H2O)and hydroxyl (OH). er than that of bulk glass as a result of the high rate of cooling dur- The mobile species, molecular water, can diffuse into and react ing manufacture. However, when the fibre is subjected to elevated with the silica to produce immobile hydroxyl by the reaction temperatures, the physical properties tend to return to those val- H20+=SH-O-SE 2=Si-OH. It is noteworthy that this ues representative of the massive, annealed form of the glass reaction breaks structural bonds, which must have consequences resulting in a related increase in the modulus. More recently, in an- on the response of the glass structure to applied stress. At high other paper from Owens Corning, Korwin-Edson et al. also reported enough temperatures, the reaction is fast and reaction equilibrium room temperature density and modulus increases in glass fibres is established with an equilibrium constant, K=[OH/H20J conditioned for one hour at temperatures from 100C to 900c where [OH] and [H2O are the concentrations of hydroxyl and 111. Their investigation covered a range of fibres made from differ- molecular water respectively [29. The concentration of dissolved ent glass formulations(including the Advantex formulation used in molecular water. [H2O. in the glass is proportional to atmospheric this study ) They reported that, after an initial dip in modulus to a water vapour pressure. Mechanical loss increases and modulus de- minimum at a conditioning temperature of 100C. the modulus of creases with increasing water content in glass 30,31 . Tomozawa eat conditioned H-glass fibre increased with increasing condition- has suggested that water enters into glasses, accelerated by tensile ng temperature due to structural changes and densification of the stress, and causes swelling which reduces modulus and strength in lass network. However they also mentioned the possible role of a manner similar to the plasticization effect in polymers [29, 30]. It urface and bulk water in the glass in connection with the glass fi- is interesting to note that the maximum values obtained by ther- mal conditioning at 600C in Fig. 5 are approximately equal to In Fig. 9 we compare our results on the change of fibre modulus the values obtained for the unconditioned fibres two years earlier as a function of conditioning temperature with those of Otto and [16]. This could imply that it is not so much that the glass fibre Korwin-Edson et al. To remove absolute differences caused by dif- modulus was increased by thermal conditioning, but that the mod- ferences in glass fibre formulation the modulus data are norma- uli of the unconditioned fibres measured in this investigation were ure unconditioned results. It is lower than expected. The thermal conditioning of the fibre then Interesting to parison to Korwin- s results, reversed this time dependent loss of modulus most likely by removal otto did not drop in fibre modulus with low of"plasticizing" water. Furthermore, the magnitude of K has been

surface. Both thermal conditioning and acid treatment have the po￾tential to significantly change surface composition, which may then change the reactivity with the silanes in comparison to the surface of pristine glass fibre. To further investigate these possibil￾ities, similar chemical treatments were applied to the unsized glass fibres without prior heat conditioning. The results for these chem￾ically treated fibres listed in Table 3 also indicate only minor ef￾fects of acid and/or silane treatment on the strength of unsized fibres that had no received thermal conditioning whatsoever. The results presented in Table 2 and 3 raise further doubts on the valid￾ity of the flaw healing theory but more importantly indicate that further work is needed to improve understanding of the role of post-silanisation in the strength of glass fibres. 3.5. Discussion of fibre modulus results Although the reduction of room temperature glass fibre strength due to an elevated temperature conditioning has been ob￾served and reported by a number of authors, there has been much less discussion of the effects on the fibre modulus. Feih et al. ob￾served similar strength losses in heat conditioned E-glass fibres as reported here. However they reported no significant change in the modulus of their fibres [10]. Kennerley et al. also reported no apparent change in the modulus of E-glass fibres recycled out of GRP at temperatures from 450 to 600 C [26]. However, Otto at Owens Corning reported that the modulus of E-glass fibre [27] at room temperature increased systematically from 75 GPa up to 86 GPa after conditioning for four hours at a temperature of 600 C. This was related to changes in the density of the glass due to densification during conditioning. Glass fibre density is low￾er than that of bulk glass as a result of the high rate of cooling dur￾ing manufacture. However, when the fibre is subjected to elevated temperatures, the physical properties tend to return to those val￾ues representative of the massive, annealed form of the glass resulting in a related increase in the modulus. More recently, in an￾other paper from Owens Corning, Korwin-Edson et al. also reported room temperature density and modulus increases in glass fibres conditioned for one hour at temperatures from 100 C to 900 C [11]. Their investigation covered a range of fibres made from differ￾ent glass formulations (including the Advantex formulation used in this study). They reported that, after an initial dip in modulus to a minimum at a conditioning temperature of 100 C, the modulus of heat conditioned H-glass fibre increased with increasing condition￾ing temperature due to structural changes and densification of the glass network. However they also mentioned the possible role of surface and bulk water in the glass in connection with the glass fi- bre modulus. In Fig. 9 we compare our results on the change of fibre modulus as a function of conditioning temperature with those of Otto and Korwin-Edson et al. To remove absolute differences caused by dif￾ferences in glass fibre formulation the modulus data are norma￾lised to the room temperature unconditioned results. It is interesting to note that, in comparison to Korwin-Edson’s results, Otto did not obtain an initial drop in fibre modulus with low temperature conditioning and recorded a much larger relative in￾crease in modulus at higher conditioning temperatures. In the case of our results it appears that the uncoated fibre results agree well with those of Otto whereas the results for the APS coated fibres ap￾pear to match those of Korwin-Edson et al. with an initial drop in modulus. In a recent paper, Yang and Thomason discussed the no￾vel use of a thermo-mechanical analyser to simultaneously mea￾sure the thermal expansion coefficient, densification, modulus of single glass fibres (the same APS coated fibres used in this investi￾gation) and also to thermally condition and measure room temper￾ature modulus in the same instrument [19]. Their results for room temperature modulus as a function of conditioning temperature (for 30 min) are also compared in Fig. 9. These data appear to show all the above discussed features with an initial drop in modulus at low conditioning temperatures and then a rapid increase in modu￾lus with conditioning above 300 C. They suggested that these observations could be explained by a combination of structural relaxation in the glass fibre combined with desorption of water from within the fibre [19]. Water can be present in significant quantities in glass and can act as plasticiser. Removal of the water from the network may stiffen the structure and contribute to the increase of the fibre modulus. Glasses manufactured using standard melt technologies always contain residual water. [28]. Water can exist in, and diffuse through, glass as both molecular water (H2O) and hydroxyl (OH). The mobile species, molecular water, can diffuse into and react with the silica to produce immobile hydroxyl by the reaction H2O + „SiAOASi„ () 2„SiAOH. It is noteworthy that this reaction breaks structural bonds, which must have consequences on the response of the glass structure to applied stress. At high enough temperatures, the reaction is fast and reaction equilibrium is established with an equilibrium constant, K = [OH]2 /[H2O], where [OH] and [H2O] are the concentrations of hydroxyl and molecular water respectively [29]. The concentration of dissolved molecular water, [H2O], in the glass is proportional to atmospheric water vapour pressure. Mechanical loss increases and modulus de￾creases with increasing water content in glass [30,31]. Tomozawa has suggested that water enters into glasses, accelerated by tensile stress, and causes swelling which reduces modulus and strength in a manner similar to the plasticization effect in polymers [29,30]. It is interesting to note that the maximum values obtained by ther￾mal conditioning at 600 C in Fig. 5 are approximately equal to the values obtained for the unconditioned fibres two years earlier [16]. This could imply that it is not so much that the glass fibre modulus was increased by thermal conditioning, but that the mod￾uli of the unconditioned fibres measured in this investigation were lower than expected. The thermal conditioning of the fibre then reversed this time dependent loss of modulus most likely by removal of ‘‘plasticizing’’ water. Furthermore, the magnitude of K has been Table 3 The effect of chemical treatments on the strength of unsized glass fibres. Sample Duration of acid treatment (h) Silane type Ave. tensile strength (GPa) 95% Confidence limit L 0 – 1.34 0.16 M 0 APS 1.55 0.17 N 0 APS:SPS 1.38 0.20 O 1 – 1.34 0.21 P 1 APS 1.52 0.16 Q 1 APS:SPS 1.56 0.14 95 100 105 110 115 0 100 200 300 400 500 600 Relative Fibre Modulus (%) Treatment Temperature (°C) Silane Sized Water Sized E-Glass (ref 27) H-Glass (ref 11) Silane Sized (ref 19) Fig. 9. Influence of heat treatment temperature on room temperature glass fibre relative modulus. 206 J.L. Thomason et al. / Composites: Part A 61 (2014) 201–208

J. L. Thomason et aL/Composites: Part A 61 (2014) 201-208 average roo te fibre tensile strength of ● Silane sized water sized and silane sized E-glass fibres was decreased by up 78 to 70% by only15 min conditioning at 600C Silane sized fibre exhibited a much higher initial strength than water sized fibres due to the protective influence of the silane surface coating. Silane sized fibres also exhibited relative stability in average tensile strength up to 250C. However, conditioning above this tempera ture resulted in a precipitous drop in room temperature strength. The water sized fibres exhibited an approximately linear decrease in average room temperature tensile strength with increasing con- d i Further analysis of the single fibre strength distributions using Weibull methods indicated that the strength distribution of the Fibre Age( Log10 Days) water sized fibres could be well represented by a unimodal, three-parameter, Weibull distribution which included the lower Fig 10. Change of glass fibre modulus with age(a water sized, APS sized) strength limit imposed by the minimum fibre strength required in order to be able to prepare a test sample. The strength distribu- tions of the sized fibres were more complicated and required the shown to increase in the temperature range 200-550C in silica use of a partially concurrent bimodal Weibull distribution. How- glass [32] implying the likelihood of increasing Si-0-Si structural ever, the use of a simple analysis of the cumulative fibre strength bond breakage with increasing conditioning temperature in the probability resulted in more useful material science understanding temperature range of this investigation. han the Weibull analysis. The observed loss in fibre strength wa Many researchers engaged in elucidating the complex struc- attributed, in part, to the thermal degradation of the organic part of an nh e relationships in composite materials operate un- the silane coating and the consequential increase in sensitivity to inforcement fibres is fixed and invariable. Consequently the the parallel changes in strength of the uncoated fibres may imply source of environmentally driven composite performance changes that there are also other changes taking place in the glass fibre is often sought in the polymer matrix or the fibre-matrix interface. structure at these relatively low processing temperatures. These results bring up the interesting possibility that long-term Fibre strength regeneration was investigated using acid surface exposure of glass fibres to an uncontrolled humid environment activation and post-silanisation of the heat-treated glass fibres. Lit- ay result in a significant reduction of fibre modulus. Since the fi- tle significant change in fibre strength was observed for a range of bres used in this investigation have also been the subject of a num- acid and silane post-treatments of heat conditioned fibres. Several the room possibile explanations were discussed including the density of su temperature modulus of these fibres over a period of approxi- face hydroxyl groups and potential changes in the fibre surface mately five years since their production. These results are shown composition. These aspects were further investigated through in Fig. 10 as a plot of average modulus versus Log(time). There silanisation of unsized glass fibres. The initial results suggest that is some scatter in the data, possibly due the uncontrolled storage recoating glass fibres with typical silanes may not be an attractive nvironment and to the large number of different investigators in- route to regenerating the value of heat conditioned glass fibres. olved in the generation of the data over this long time period. More work is required to further investigate the potential of Nevertheless, there is a clear trend visible in the data showing a post-silanisation in strength regeneration of thermally recycled time dependent lowering of the fibre modulus of both the water glass fibres. In contrast to the fibre strength, the fibre modulus sized and silane sized glass fibres. Fig. 10 also shows the expected was found to increase with increasing thermal conditioning tem- value of fibre modulus for these samples after 10 and 25 years. At perature. This was discussed in terms of the possible effects of ab- 25 years the fibre modulus could be below 62 GPa, which is more sorbed water on fibre properties. It was found that the modulus of than 20% reduction from the initial modulus of an Advantex glass these fibre decrease logarithmically as a function of age. It was sug- fibre. If this phenomenon is associated with slow moisture diffu- gested observed that this effect could be of significance for the sion into the glass fibres then clearly one might expect an acceler- long-term performance of glass fibres reinforced composites. ation of the effect at higher temperature and humidity. Consequently, we consider that this effect could be of great impor- tance for the long term performance of glass fibres (and their com- Acknowledgements posites)in humid environments such as offshore wind turbine blades, or the short term performance of glass fibres(and compos The authors would like to thank Owens Corning-Vetrotex for es)that experience hot-wet conditions such as composites used providing the glass fibres used in this study. The authors would in automotive cooling systems. Clearly these considerations merit also like to acknowledge the financial support from Engineering further controlled investigations and we have initiated a dedicated and Physical Sciences Research Council under the Recover and study of time-dependent changes in the modulus of E-glass fibres TARF-LCv projects and the assistance of the Advanced Materials stored in controlled environments. This will be reported at a later Research Laboratory of the University of Strathclyde with the mechanical testing. 4. Conclusions References The results of single fibre tensile testing presented here clearly show that the thermal history likely to be encountered by glass fi- [1I Job S Recycling glass fibre reinforced composites- history and progress. res recycled out of end-of-life composit 250-600C temperature range can potentially cause dramatic low- Composites in the InterAgency Composites Group Department for ovation ering of the room temperature fibre strength and strain at failure November 2009.

shown to increase in the temperature range 200–550 C in silica glass [32] implying the likelihood of increasing SiAOASi structural bond breakage with increasing conditioning temperature in the temperature range of this investigation. Many researchers engaged in elucidating the complex struc￾ture-performance relationships in composite materials operate un￾der an initial assumption that the modulus of inorganic reinforcement fibres is fixed and invariable. Consequently the source of environmentally driven composite performance changes is often sought in the polymer matrix or the fibre–matrix interface. These results bring up the interesting possibility that long-term exposure of glass fibres to an uncontrolled humid environment may result in a significant reduction of fibre modulus. Since the fi- bres used in this investigation have also been the subject of a num￾ber of other projects it is possible to examine the room temperature modulus of these fibres over a period of approxi￾mately five years since their production. These results are shown in Fig. 10 as a plot of average modulus versus Log (time). There is some scatter in the data, possibly due the uncontrolled storage environment and to the large number of different investigators in￾volved in the generation of the data over this long time period. Nevertheless, there is a clear trend visible in the data showing a time dependent lowering of the fibre modulus of both the water sized and silane sized glass fibres. Fig. 10 also shows the expected value of fibre modulus for these samples after 10 and 25 years. At 25 years the fibre modulus could be below 62 GPa, which is more than 20% reduction from the initial modulus of an Advantex glass fibre. If this phenomenon is associated with slow moisture diffu￾sion into the glass fibres then clearly one might expect an acceler￾ation of the effect at higher temperature and humidity. Consequently, we consider that this effect could be of great impor￾tance for the long term performance of glass fibres (and their com￾posites) in humid environments such as offshore wind turbine blades, or the short term performance of glass fibres (and compos￾ites) that experience hot-wet conditions such as composites used in automotive cooling systems. Clearly these considerations merit further controlled investigations and we have initiated a dedicated study of time-dependent changes in the modulus of E-glass fibres stored in controlled environments. This will be reported at a later stage. 4. Conclusions The results of single fibre tensile testing presented here clearly show that the thermal history likely to be encountered by glass fi- bres recycled out of end-of-life composite by processing in the 250–600 C temperature range can potentially cause dramatic low￾ering of the room temperature fibre strength and strain at failure. The average room temperature single fibre tensile strength of water sized and silane sized E-glass fibres was decreased by up to 70% by only15 min conditioning at 600 C. Silane sized fibres exhibited a much higher initial strength than water sized fibres due to the protective influence of the silane surface coating. Silane sized fibres also exhibited relative stability in average tensile strength up to 250 C. However, conditioning above this tempera￾ture resulted in a precipitous drop in room temperature strength. The water sized fibres exhibited an approximately linear decrease in average room temperature tensile strength with increasing con￾ditioning temperature. Further analysis of the single fibre strength distributions using Weibull methods indicated that the strength distribution of the water sized fibres could be well represented by a unimodal, three-parameter, Weibull distribution which included the lower strength limit imposed by the minimum fibre strength required in order to be able to prepare a test sample. The strength distribu￾tions of the sized fibres were more complicated and required the use of a partially concurrent bimodal Weibull distribution. How￾ever, the use of a simple analysis of the cumulative fibre strength probability resulted in more useful material science understanding than the Weibull analysis. The observed loss in fibre strength was attributed, in part, to the thermal degradation of the organic part of the silane coating and the consequential increase in sensitivity to fibre surface mechanical and environmental damage. Nevertheless, the parallel changes in strength of the uncoated fibres may imply that there are also other changes taking place in the glass fibre structure at these relatively low processing temperatures. Fibre strength regeneration was investigated using acid surface activation and post-silanisation of the heat-treated glass fibres. Lit￾tle significant change in fibre strength was observed for a range of acid and silane post-treatments of heat conditioned fibres. Several possibile explanations were discussed including the density of sur￾face hydroxyl groups and potential changes in the fibre surface composition. These aspects were further investigated through silanisation of unsized glass fibres. The initial results suggest that recoating glass fibres with typical silanes may not be an attractive route to regenerating the value of heat conditioned glass fibres. More work is required to further investigate the potential of post-silanisation in strength regeneration of thermally recycled glass fibres. In contrast to the fibre strength, the fibre modulus was found to increase with increasing thermal conditioning tem￾perature. This was discussed in terms of the possible effects of ab￾sorbed water on fibre properties. It was found that the modulus of these fibre decrease logarithmically as a function of age. It was sug￾gested observed that this effect could be of significance for the long-term performance of glass fibres reinforced composites. Acknowledgements The authors would like to thank Owens Corning-Vetrotex for providing the glass fibres used in this study. The authors would also like to acknowledge the financial support from Engineering and Physical Sciences Research Council under the ReCoVeR and TARF-LCV projects and the assistance of the Advanced Materials Research Laboratory of the University of Strathclyde with the mechanical testing. References [1] Job S. Recycling glass fibre reinforced composites – history and progress. Reinf Plast 2013;57(5):19–23. [2] Technology Needs To Support Advanced Composites in the UK. Written by The InterAgency Composites Group. Department for Business, Innovation & Skills. November 2009. . 60 62 64 66 68 70 72 74 76 78 80 82 2.0 2.5 3.0 3.5 4.0 Modulus (GPa) Fibre Age (Log10 Days) Silane Sized Water Sized 10yr extrapolated 25yr extrapolated Fig. 10. Change of glass fibre modulus with age (N water sized, d APS sized). J.L. Thomason et al. / Composites: Part A 61 (2014) 201–208 207

J.L. Thomason et aL/Composites: Part A 61 (2014)201-208 [3 Red C wind turbine blades: big and getting bigger. Composites world ju [191 Yang L, Thomason JL The thermal behaviour of glass fibre investigated by 2008.http://www.compositesworld.com/articles/wind-turbine-blades-big- thermomechanical analysis. J Mater Sci 2013: 48: 5768-75 A statistical distribution function of wide applicability. J Appl [4 Larsen K Recycling wind Reinf Plast 2009: 53: 20-5. Aech1951:18:293-7. Williams PT, Cunliffe A Jones N Recovery of value-added products from the [21 Zinck P, Pay MF, Rezakhanlou R, Gerard JF. Mechanical characterisation of glass bres as an indirect analysis of the effect of surface treatment. Mater Sc [6] oliveux G, Bailleul JL, Salle ELGL Chemical recy 1999:34:2121-33 I Feih S, Thraner A, Lilholt H. Tensile strength and fracture surface rent status. Compos A 2006: 37: 1206- ai9206 [9 Feih S, Manatpon K, Mathys Z Gibson AG, Mouritz AP Strength degradation of [24 Gonzalez-Benito J Baselga J. Aznar AJ Microst 101d urface pretreated glass fibres. J Mater Process Technol 1999: 92-93: 129-34. Y, Tavman IH, Erkan G, Cecen V. The structure of y- thermally-treated and recycled glass fibres Compos B 2011: 42(3): 350-8. glycidoxypropyltrimethoxysilane on glass fiber surfaces: characterization by [111 Korwin-Edson ML, Hofmann DS, McGinnis PB Strength of high performance FTIR. SEM. and angle measurements. Polym Compos 2009 glass reinforcement fiber. Int J Appl Glass Sci 2012: 3(2): 107-21 [12] Thomason JL, Kao CC, Ure ], Yang L The strength of glass fibre reinforcement [26] Kennerley JR, Kelly RM, Pickering S). Rudd CD, Characterisation and reuse of 20+915pM es recycled from scrap composites by a fluidised bed process. Compos 199829:839-45 [13] Piggott MR, Yokom JC. The weakening of silica fibres by heat treatment. Glass [27 Otto WH. Compaction effects in glass fibers. J Am Ceram Soc 1961 echnol1968:9:172- [14] Militky J Kovacic V, Rubnerova ). Influence of thermal treatment on tensile [281 Naraev VN The influence of water on the glass properties. Glass Phys Chem failure of basalt fibers. Eng Fract Mech 2002: 69: 1025-33 2004; [16 Yang L Thomason JL Effect of silane coupling agent on mechanical erformance of glass fibre. J Mater Sci 2013: 48: 1947-54. [301 Ito S, Tomozawa M. Dynamic fatigue of sodium-silicate glasses with high [17 Thomason JL On the application of weibull analysis to experimentally issolved water on the internal friction of glass. J Non-Cryst Solids 1974: 14(1): 165-77 [18 Jenkins PG, Yang L, Liggat I, Thomason JL. Investigation of the strength loss of E- [32] Tomozawa M. Kim DL, Agarwal A, Davis KM. Water diffusion and surface lass fibres after thermal conditioning. J Mater Sci [submitted for publication- structural relaxation of silica glasses. J Non-Cryst Solids 2001: 288: 73-80

[3] Red C. Wind turbine blades: big and getting bigger. Composites World June 2008. . [4] Larsen K. Recycling wind. Reinf Plast 2009;53:20–5. [5] Williams PT, Cunliffe A, Jones N. Recovery of value-added products from the pyrolytic recycling of glass-fibre composite waste. J Energy Inst 2005;78:51–61. [6] Oliveux G, Bailleul JL, Salle ELGL. Chemical recycling of glass fibre reinforced composites using subcritical water. Compos A 2012;43(11):1809–18. [7] Pickering SJ. Recycling technologies for thermoset composite materials￾current status. Compos A 2006;37:1206–15. [8] Thomas WF. An investigation of the factors likely to affect the strength and properties of glass fibres. Phys Chem Glasses 1960;1:4–18. [9] Feih S, Manatpon K, Mathys Z, Gibson AG, Mouritz AP. Strength degradation of glass fibers at high temperatures. J Mater Sci 2009;44:392–400. [10] Feih S, Boiocchi E, Mathys G, Gibson AG, Mouritz AP. Mechanical properties of thermally-treated and recycled glass fibres. Compos B 2011;42(3):350–8. [11] Korwin-Edson ML, Hofmann DS, McGinnis PB. Strength of high performance glass reinforcement fiber. Int J Appl Glass Sci 2012;3(2):107–21. [12] Thomason JL, Kao CC, Ure J, Yang L. The strength of glass fibre reinforcement after exposure to elevated composite processing temperatures. J Mater Sci 2014;49(1):153–62. [13] Piggott MR, Yokom JC. The weakening of silica fibres by heat treatment. Glass Technol 1968;9:172–5. [14] Militky´ J, Kovacˇicˇ V, Rubnerová J. Influence of thermal treatment on tensile failure of basalt fibers. Eng Fract Mech 2002;69:1025–33. [15] Thomason JL. The influence of fibre properties of the performance of glass- fibre-reinforced polyamide 6,6. Compos Sci Technol 1999;59(16):2315–28. [16] Yang L, Thomason JL. Effect of silane coupling agent on mechanical performance of glass fibre. J Mater Sci 2013;48:1947–54. [17] Thomason JL. On the application of weibull analysis to experimentally determined single fibre strength distributions. Compos Sci Technol 2013;77:74–80. [18] Jenkins PG, Yang L, Liggat JJ, Thomason JL. Investigation of the strength loss of E￾glass fibres after thermal conditioning. J Mater Sci [submitted for publication]. [19] Yang L, Thomason JL. The thermal behaviour of glass fibre investigated by thermomechanical analysis. J Mater Sci 2013;48:5768–75. [20] Weibull W. A statistical distribution function of wide applicability. J Appl Mech 1951;18:293–7. [21] Zinck P, Pay MF, Rezakhanlou R, Gerard JF. Mechanical characterisation of glass fibres as an indirect analysis of the effect of surface treatment. J Mater Sci 1999;34:2121–33. [22] Feih S, Thraner A, Lilholt H. Tensile strength and fracture surface characterisation of sized and unsized glass fibers. J Mater Sci 2005; 40:1615–23. [23] Zhuravlev LT. The surface chemistry of amorphous silica. Colloids Surf A 2000;173:1–38. [24] González-Benito J, Baselga J, Aznar AJ. Microstructural and wettability study of surface pretreated glass fibres. J Mater Process Technol 1999;92–93:129–34. [25] Sever K, Seki Y, Tavman IH, Erkan G, Cecen V. The structure of c￾glycidoxypropyltrimethoxysilane on glass fiber surfaces: characterization by FTIR, SEM, and contact angle measurements. Polym Compos 2009; 30(5):550–8. [26] Kennerley JR, Kelly RM, Pickering SJ, Rudd CD. Characterisation and reuse of glass fibres recycled from scrap composites by a fluidised bed process. Compos A 1998;29:839–45. [27] Otto WH. Compaction effects in glass fibers. J Am Ceram Soc 1961; 44(2):68–72. [28] Naraev VN. The influence of water on the glass properties. Glass Phys Chem 2004;30(5):367–89. [29] Lezzi PJ, Xiao QR, Tomozawa M, Blanchet TA, Kurkjian CR. Strength increase of silica glass fibers by surface stress relaxation. J Non-Cryst Solids 2013; 379:95–106. [30] Ito S, Tomozawa M. Dynamic fatigue of sodium-silicate glasses with high water content. J Phys 1982;C9:611–4. [31] Day DE, Stevels JM. Effect of dissolved water on the internal friction of glass. J Non-Cryst Solids 1974;14(1):165–77. [32] Tomozawa M, Kim DL, Agarwal A, Davis KM. Water diffusion and surface structural relaxation of silica glasses. J Non-Cryst Solids 2001;288:73–80. 208 J.L. Thomason et al. / Composites: Part A 61 (2014) 201–208