Improving Impact Value of Interlayered Glass Fiber Chopped Strand Mat Reinforced Thermoplastic Polypropylene Externally Irradiated by Homogeneous Low Potential Electron Beam
2016 Volume 57 Issue 11 Pages 1915-1921
Details
2016 Volume 57 Issue 11 Pages 1915-1921
We propose a new method of applying homogeneous low potential electron beam irradiation (HLEBI) externally to both sides of glass fiber (GF) chopped strand mat (GF-CSM) reinforced thermoplastic polypropylene (PP) interlayered samples (CSM-GFRTP) after lamination assembly and hot press with layup of [PP]6[GF-CSM]5. The external process apparently improved the Charpy impact values (auc) over that of untreated. HLEBI dose from 0.22 to 0.43 MGy apparently enhanced the auc at low cumulative probability of fracture (Pf = 0.06) 40% over the untreated from 48 to 67 kJ/m2. Moreover, based on the 3-parameter Weibull equation, applying 0.30 MGy HLEBI increased the statistically lowest impact value, as at Pf = 0 significantly 120% over that of the untreated from 29 to 64 kJm−2. The improvements of weakest samples in the data set indicate increase in reliability and safety. The improvement in auc can be attributed to HLEBI generating dangling bonds in the PP polymer evidenced by an ESR peak with inflection point at B = ～322.5 mT. The HLEBI apparently acts to generate nano-compressive stresses from repulsive forces between the severed bond electrons in the PP and GFs hence strengthening the GF and PP as well as GF/PP interface causing rise in impact energy. Therefore, the external HLEBI application aims to be a viable method to increase Charpy impact value in articles of GF-CSM reinforced PP (CSM-GFRTP) for industry such as automotive, aerospace and construction.
Glass fiber reinforced polymers (GFRPs) are strong, lightweight materials that have been used in numerous daily articles for over a half century. GF reinforcement has many forms such as short fibers, layed up long fibers, filament winded, pultruded, or weaves. A widely used GF reinforcement is chopped strand mat (CSM) where fibers typically 50 to 60 mm in length are layed randomly into a flat mat giving in-plane isotropic properties. Typically referred to as “fiberglass”, a famous application for CSM is the Corvette body. Polymer matrix used with CSM are typically thermoset epoxies or vinyl esters (CSM-GFRPs), and less frequently thermoplastics (CSM-GFRTPs). GFRP-CSMs have wide application in aerospace, automotive, construction and sports industries such as: aircraft, car bodies, boats, bathtubs, swimming pools, roofing, pipes and surfboards. The randomly placed long fibers with high volume fraction act to prevent matrix microcrack propagation over the wide area of the composite structure and give it rigidity.
Thermoplastics have the advantages of having higher crack resistance, easier formability, and can be more easily recycled than thermosets although they have lower ultimate strength. Despite lower strength, they can be used for many applications. Therefore, we suggest a new method for strengthening GFRTP-CSM.
It is always important to increase strength of composite parts for increased safety, reliability and lifetime. To do this, one method that has been gaining momentum is low potential 100-keV class electron beam irradiation (HLEBI).1–8) HLEBI has a proven track record of increasing desirable properties such as hardening, wear resistance, mist resistance and sterilization for practical use of polymer.1–3) A small dose of HLEBI often enhances deformation resistivity (elasticity) of silicate glasses.4)
Moreover, HLEBI has improved the bending strength of CFRP,6) as well as bending fracture energy of GFRP.7) Our recent research has shown HLEBI enhances Charpy impact strength of thermohardened epoxy matrix-GFRP.8)
Thermoplastic GFRTPs, notably, already applied for automobile bumper and IC (integrated circuit) substrates, can be easily produced with high productivity for mass production by various forming process. One of the typical thermoplastic polymers is polypropylene (PP), which is constructed of hydrogen and carbon. Since its production cycle is short enough to advance cost reduction, it can be generally applied to automobiles. However, the typical poor adhesion of thermoplastic polymer with glass fibers has been a serious problem.
To remedy this, our previous study demonstrated internal HLEBI activation throughout the bulk of CSM-GFRTP sample thickness increased Charpy impact values, especially the weakest samples in the data sets (low fracture probabilities, Pf < 0.10) significantly over those of untreated.9) Each CSM ply of interlayered composite was irradiated by HLEBI prior to lamination assembly and hot press to couple with difficult to adhere thermoplastic PP.9) This was for the layup of thermoplastic [PP/GF/PP/GF/PP/GF/PP/GF/PP/GF/PP] referred to here as [PP]6[GF-CSM]5. The purpose was to maximize the fiber surface area (fiber diameter: df = ～10 μm) activated by HLEBI throughout the bulk thickness. Results shown Charpy impact value9) were improved over untreated using internal method. Moreover, this shows theoretically CSM-GFRTPs samples with infinite thickness can be prepared by an internal HLEBI activation to GF prior to dipping into PP. However, a disadvantage of internal HLEBI activation is it requires an additional N − 1 extra steps in Step 1 applying the HLEBI to each GF-CSM ply as shown in Fig. 1, bottom. Moreover, internal activation cannot be done after molding and shaping: in this case external HLEBI activation is required.
It follows external HLEBI application to samples has been successfully developed.5–8,10–16) Conventional method of HLEBI to outside surfaces after molding and shaping has resulted in improvements and high reproducibly of mechanical properties in both GF and polymer, as well as enhancing fiber/polymer interface adhesion. This is despite HLEBI penetration depth, Dth being on the order of about 166 μm into specimen thickness for typical GFRP and CFRPs. External HLEBI activation is simpler than the internal having less steps as shown in the top of Fig. 1 which saves cost and time.
Therefore, although internal activation has its suitable applications, to be cost and time effective we propose the conventional simpler process10–16) of external activation applying HLEBI to the outer surfaces of the finished [PP]6[GF-CSM]5 CSM-GFRTP samples after lamination assembly and hot press.
Since Charpy impact is a surface sensitive test, we predict the new method of external HLEBI activation as well to the [PP]6[GF-CSM]5 CSM-GFRTP will increase Charpy impact strength significantly over the untreated especially at low Pf < 0.10 while increasing statistical reliability about as much as that of the internal HLEBI activation. When applying to actual parts, careful consideration to adjusting HLEBI to get optimum dose is highly recommended for reliability and safety.
GFRTP composite samples with volume of 2000 mm3 have been constructed with glass fiber (GF) chopped strand mat (CSM) (MC 300 A-104CM Nittobo) interlayered with thermoplastic polymer (Polypropylene, PP; BC06C D23510 Novatec) matrix. Individual PP sheets were ～0.5 mm thickness. In the CSM, the GF dimensions were length, lfiber = ～60 mm with nominal fiber diameter, df = 10 μm. Regarded as short fibers (compared with unidirectional which typically span the entire panel) fibers were at random angles in the mat. Samples were prepared as follows: Step 1: The composite was assembled consisting of 5 GF-CSM mats assembled between 6 sheets of PP. Step 2: Panels were heated (laminated) by hot press at 473 K for 9 min under 2.0 MPa. Step 3: Individual samples were cut. Step 4: The individual finished samples were irradiated with HLEBI on both side surfaces. Volume fractions, Vf of GF and PP matrix were 44.5 and 55.5 vol%, respectively.
2.2 Charpy impact testIn order to evaluate the impact fracture toughness, the Charpy impact values (auc) of the CSM-GFRTP with and without HLEBI were measured using a standard impact fracture energy measurement system (Shimadzu Corporation No.51735) (JIS K 7077).11,17) Sample dimensions were: [length × width × thickness] = [80 × 10 × 2.5 mm]. The impact value is expressed by the following equation.11,17)
| \[E = WR[(\cos\beta - \cos\alpha) - (\cos\alpha' - \cos\alpha)(\alpha + \beta)/(\alpha - \alpha')]\] | (1) |
| \[a_{uc} = E/(b \times t)\] | (2) |
An electron-curtain processor (Type CB175/15/180L, Energy Science Inc., Woburn, MA, Iwasaki Electric Group Co., Ltd., Tokyo) was used to irradiate the samples homogeneously18–23) by linear electron beam gun with low energy through a titanium thin film window attached to a 240 mm diameter vacuum chamber. As shown in Fig. 2, a tungsten filament in a vacuum was used to generate the electron beam at a low energy (acceleration potential, V: kV), of 170 kV and irradiating current density (I, A/m2) of 0.089 A/m2.
Schematic diagram of electron curtain processer. (Iwasaki Electric Group Co., Ltd., Tokyo)
Although the electron beam was generated in a vacuum, the irradiated sample was kept in inert condition under protective nitrogen at atmospheric pressure. The distance between sample and window was 25 mm. To prevent oxidation, the samples were kept in a protective atmosphere of nitrogen gas with a residual concentration of oxygen below 300 ppm. The flow rate of nitrogen gas was 1.5 L/s at 0.1 MPa nitrogen gas pressure. The sample in the aluminum plate holder (0.15 m × 0.15 m) was transported on a conveyor at a speed of 10 m/min. The sheet electron beam irradiation was applied intermittently to avoid excessive heating of the sample; the temperature of the sample surface remained below 323 K just after irradiation, one sweep going one way was 43.2 kGy. Repeated irradiations to both side surfaces of the samples were used to increase the total irradiation dose. The interval between the end of one period of irradiation and the start of the next operation was 30 s. When the irradiation current (I, mA), the conveyor speed (S, m/min) and number of irradiations (N) are determined, the irradiation dose is expressed by the following equation:
| \[\text{Dosage (MGy)} = 0.216 \times ({\rm I/S}) \times {\rm N}\] | (3) |
The irradiation dose was controlled by the integrated irradiation time in each of the samples. Here, irradiation dose was corrected by using an FWT nylon dosimeter of RCD radiometer film (FWT-60-00: Far West Technology, Inc. 330-D South Kellogg Goleta, California 93117, USA) with an irradiation reader (FWT-92D: Far West Technology, Inc. 330-D South Kellogg Goleta, California 93117, USA).
Based on the mean density of the samples (ρ: kg/m3) and irradiation potential at the sample surface (V: keV), the penetration depth (Dth:/m) of HLEBI is expressed by the following equation.24)
| \[D_{\rm th} = 66.7V^{5/3}/\rho\] | (4) |
| \[V = 170\,{\rm keV} - \mathit{\Delta} V_{\rm Ti} - \mathit{\Delta} V_{\rm N2}\] | (5) |
| \begin{align*} \mathit{\Delta} V_{\rm Ti} &= T_{\rm Ti}/D_{\rm thTi} \times 170\,{\rm keV} = T_{\rm Ti} \rho_{\rm Ti}/[ 66.7 \times (170\,{\rm keV})^{5/3}] \\ &= (10 \times 10^{-6}\,{\rm m}) \times (4540\,{\rm kgm}^{-3})/[66.7 \times (170\,{\rm keV})^{5/3}] \\ & = 22.2\,{\rm keV} \end{align*} | (6) |
| \begin{align*} \mathit{\Delta} V_{\rm N2} & = T_{\rm N2}/{\rm D}_{\rm thN2} \times V_{\rm Ti} = T_{\rm N2} \rho _{\rm N2}/[66.7 \times (V_{\rm Ti})^{5/3}] \\ &= (35 \times 10^{-3}\,{\rm m}) \times (1.13\,{\rm kgm}^{-3})/[66.7 \times (170\,{\rm keV})^{5/3}] \\ &= 18.2\,{\rm keV} \end{align*} | (7) |
| \[V = 170\,{\rm keV} - 22.2\,{\rm keV} - 18.2\,{\rm keV} = 129.6\,{\rm keV}\] | (8) |
Since typical densities of PP and GF are 900 and 2620 (E-Glass Fibers) kgm−3, the irradiation depth, Dth into the PP and GF estimated from eq. (4) are about 247 and 84 μm, respectively. Since the volume fractions Vf of GF and PP matrix are 44.5 and 55.5 vol%, estimated Dth of the CSM-GFRTP is ～157 μm (about 6.3% into the 2.5 mm thickness, total 12.6% at both side surfaces) of the CSM-GFRTP. Based on the experimental results of ESR signals15), the experimental Dth of the PEEK polymer is more than 300 μm. The experimental irradiation depth of the CSM-GFRTP can be predicted to be more than 180 μm (about 7.2% into the 2.5 mm thickness, total 14.4% at both side surfaces) of the CSM-GFRTP. Based on the estimated and semi-experimental data, the effective irradiation depth is probably from 150 to 200 μm of the CSM-GFRTP.
2.4 Evaluation of dangling bondsTo obtain more precise information on atomic-scale structural changes in the PP, the dangling bond density was obtained using an electron spin resonance spectrometer (ESR, JES-FA2000, Nippon Denshi Ltd. Tokyo).25) ESR is utilized to detect unpaired electrons by their spins (ms = +/−1/2) since electrons have a magnetic moment and spin quantum number. The unpaired electrons' magnetic moments either align themselves parallel or anti-parallel to an induced magnetic field producing a peak at a particular magnetic field, B. The microwave frequency range used in the ESR analysis was the X-band at 9.45 ± 0.05 GHz with a field modulation of 100 kHz. The microwave power was 1 mW. Magnetic field, B was varied from 317 to 327 mT. The spin density was calculated using a Mn2+ standard sample. Only ESR spectra, instead of spin densities, were given since relative peak intensities conveniently compare number of dangling bonds (spin density) generated in the samples from the HLEBI.
Evaluating the accumulative probability of fracture (Pf) is a convenient method of quantitatively analyzing experimental values relating to fracture. Pf is expressed by the following equation, which is a generalized form of the median rank method:26)
| \[P_{\rm f} = (I - 0.3)/(N_s + 0.4)\] | (9) |
Figure 3 shows change in impact value (auc) of the [PP]6[GF-CSM]5 CSM-CFRTP against fracture probability at each HLEBI dose. HLEBI apparently improved the auc at accumulative probability of fracture (Pf) from 0.4 to 0.7. Figure 3 also shows HLEBI dose from 0.13 to 0.65 MGy improves the auc of at each Pf up to 0.68.
Relationships between Charpy impact value (auc) of CSM-CFRTP and accumulative probability of fracture (Pf) at each HLEBI dose.
Figure 4 shows changes in auc at low-, median- and high Pf of 0.06, 0.50 and 0.94 along with the as, the statistically lowest impact value estimated, auc at Pf = 0 (see Section 4.1) against HLEBI dose. At median-Pf of 0.50 all HLEBI levels tested from 0.04 to 0.65 MGy slightly enhance the auc of the GFRTP. Although Figs. 3 and 4 show at high-Pf of 0.94 the irradiation does not increase the auc, Fig. 4 shows remarkable effects of HLEBI from 0.13 to 0.65 MGy on the auc have been obtained at low-Pf of 0.06. Specifically, HLEBI dose from 0.22 to 0.43 MGy apparently enhances the auc at Pf = 0.06, 37% from 48 to 65–67 kJ/m2. Hence, since the weakest samples are improved, the 0.22 to 0.43 MGy HLEBI increases reliability (reproducibly) and safety while decreasing experimental errors.
Changes in experimental impact values of external HLEBI activation (solid lines) at low-, median- and high- accumulative fracture probability (0.06, 0.50 and 0.94) along with iterated as at Pf = 0, together with reported data for internal9) HLEBI activation (dotted lines) of the [PP]6[GF-CSM]5 layup CSM-GFRTP samples.
HLEBI generally produces detectable dangling bonds in thermoplastic Polyetheretherketone (PEEK), Phenol and silica glass.12,13,27) To discuss the influences of electron beam irradiation on the Charpy impact values, ESR signals related to dangling bonds have been investigated. Figure 5 shows the ESR signals of polypropylene (PP) sheet with and without HLEBI. Although ESR signals of the untreated PP sample are not detected, ESR signals are observed for the irradiated PP sheet corresponding to dangling bonds. Based on the standard calibration material TEMPOL (2,2,6,6-tetramethyl -4-piperidinol-1-oxyl) and Mn2+ in the MnO, the dangling bond density is estimated by the double integrated intensity of ESR signal.22,23) As shown in Fig. 5, an ESR peak is generated with an inflection point at magnetic field B = ～322.5 mT whose intensity increases with HLEBI dose appearing to reach a maximum at 0.22 or 0.30 MGy. Subsequently, from 0.30 to 0.65 MGy, ESR peak intensity appears to level off. This may explain the maxima in auc of the weakest samples of low-Pf = 0.06 from 0.22 to 0.43 MGy in Fig. 4 showing dangling bond generation successfully increases reliability of impact strength in the CSM-CFRTP.
ESR signals of pure PP sheet before and after HLEBI.
In order to obtain the statistically lowest impact value for safety design often used in quality control (QC), the lowest auc value at Pf = 0 (as) is assumed to be attained from the adaptable relationship of the 3-parameter Weibull equation iterating to the high correlation coefficient (F). The Pf depends on the risk of rupture ([auc − as]/aIII).13–16,27–29)
| \[P_{\rm f} = 1 - \exp [-([a_{\rm uc} - a_{\rm s}]/a_{\rm III})^m]\] | (10) |
| \[\ln[-\ln(1 - P_{\rm f})] = m \ln[(a_{\rm uc} - a_{\rm s})/a_{\rm III}]\] | (11) |
In predicting the as, coefficient (m) and constant (aIII) are the key parameters. When the term ln[−ln(1 − Pf)] is zero, Pf is 0.632 and (auc − as) = aIII. The aIII value is determined, when the auc value at Pf = 0.632 (auc (0.632)) is equal to (aIII + as) value. When Pf = 0, the required auc value to evaluate new structural materials is defined as the as. Figure 6 shows the results of the iteration to obtain the highest correlation coefficient (F) with respect to the potential auc value (eauc) estimated from the logarithmic form eq. (11).
Changes in correlation coefficient (F) against the potential as value (eas) of CSM-GFRTP irradiated by HLEBI at each dosage. The lowest impact value (as) at Pf = 0 is determined at the maximum F value.
Plots of the as with the experimental auc values at low-, median- and high-Pf. are shown in Fig. 4. The as is always lower than the experimental auc value at low-Pf = 0.06. Figure 4 shows applying the 0.30 MGy increases the as at Pf = 0 significantly 120% over that of the untreated from 29 to 64 kJm−2 indicating a decent increase in reliability and safety. Moreover, Figs. 4 and 6 show applying HLEBI of 0.13. 0.22, 0.43 and 0.65 MGy increases the as over the untreated (29 kJm−2) to 35, 62, 52 and 48 kJm−2, respectively. Figures 4 and 6 also show from 0.22 to 0.43 MGy, the as appears to exhibit its highest values over 50 kJ/m2.
Figure 7 shows the linear relationships between ln([auc − as]/aIII) and ln[−ln(1 − Pf)]. The values of aIIIand m are determined by the least-squares best fit method. The m value is estimated by the slope of the relationship when eauc = as.
Linear relationships between ln[(auc − as)/aIII] and ln[−ln(1 − Pf)] of CSM-GFRTP irradiated by HLEBI at each dosage.
Based on the ESR signals in Fig. 5, dangling bonds are not detected in pure PP sheet before treatment. On the other hand, applying the small HLEBI dose enhances the density of dangling bonds, and then increases mechanical properties in inorganic glass4,6,7) and often enhances ductility and strengthening of polymer matrix9,10) and other GFRPs.5,9–11)
In addition, HLEBI is also reported to generate dangling bonds in silica glass.27,30) Since small dose of HLEBI generates dangling bonds in the silica glass27), partial relaxation of residual strain probably occurs around these dangling bonds in the network structure. If this relaxation resulted in optimization of the interatomic distance of the silicon–oxygen pairs to minimize the potential energy, it would increase the bonding energy of the network structure. The increased impact value was therefore mainly due to an increase in the bonding energy for the silicon–oxygen atomic pairs in the atomic network structure, as well as the relaxation of the network structure.
Furthermore, when HLEBI irradiation generates active terminated atoms with dangling bonds, repulsive force is created between the outer shell electrons of the PP matrix and GF. These repulsive forces act as nano expansion sites, which like the role of a cushion with compressive stress and strain in PP and GF, as well as the GF/PP interfacial friction force, mainly resulting in strengthening impact resistance of the CSM-GFRTP.
Dangling bonds are typically created by HLEBI between atoms with low dissociation energies. Dissociation energies of groups in PP are reported to be 369 kJmol−1 for CH2-CH2; and 427 kJmol−1 for H-CH31,32) which are considered low level energies where dangling bonds are reported to form by HLEBI in other polymers.5) Moreover, the 369 and 427 kJmol−1 bond dissociation energies in PP are apparently low enough to generate ESR signals in Fig. 5.
When HLEBI irradiation generates active terminated atoms with dangling bonds at each surface of PP and GF, the strengthening of fiber pull-out induced by adhesion also occurs at GF/Polymer interface5,9), partly resulting in improving impact resistance of CSM-GFRTP.
On the other hand, Fig. 4 shows Charpy impact values, auc and the statistically lowest as at Pf = 0 begin to drop off when higher HLEBI irradiation levels of 0.43 and 0.65 MGy are applied. An excess of dangling bond density, i.e. dissociated bonds probably accelerate crack initiation and propagation rates. Therefore, carefulness is highly recommended to optimize HLEBI dose for maximum reliability and safety for CSM GFRTP parts in industry.
4.3 HLEBI-strengthening GFRP surface compared with strengthening bulk by HLEBI-activation of GFs prior to dipping in the PPAs shown in Fig. 1 and Fig. 4, external5–8,10–16) and internal9) HLEBI activation fabrication processes of [PP]6[GF-CSM]5 CSM-CFRTP are different. Since strengthening GFRP bulk has been suggested and developed to reinforce the GF/PP interface adhesion by the internal HLEBI activation of GFs prior to dipping in the PP (see broken lines in Fig. 4)9), GFRP bulk sheets with infinite thickness can be prepared. However, this process cannot be effective after molding and shaping.
On the other hand, the external HLEBI activation strengthens the GFRTP surface after molding and shaping with high reproducibly12,13,15,16) has been successfully developed to reinforce the outside of about 100 μm depth of fibers and polymer, as well as GF/PP interface adhesion (see solid lines in Fig. 4).
The maximum impact value with high reproducibly are found at large and small dose of 0.63 MGy for GFRP bulk by the internal HLEBI activation to GFs prior to dipping in the PP and 0.30 MGy for GFRTP surface activated by the external HLEBI after molding and shaping, respectively. Thus, the 2nd important point to reduce the cost is that the HLEBI-dose of external activated GFRTP surface irradiated (solid lines in Fig. 4) is half of that of GFRP bulk by internal HLEBI activation to GFs prior to dipping in the PP (broken lines in Fig. 4).
In summary, applying homogeneous low potential electron beam irradiation (HLEBI) externally to both sides of glass fiber reinforced thermoplastic (GFRTP) polypropylene (PP) chopped strand mat (CSM) samples after lamination assembly and hot press apparently improved the Charpy impact values (auc) over that of untreated. The sample layup was [PP/GF/PP/GF/PP/GF/PP/GF/PP/GF/PP].
(1) HLEBI dose from 0.22 to 0.43 MGy apparently enhanced the auc at low accumulative fracture probability (Pf = 0.06) 40% over the untreated from 48 to 67 kJ/m2.
(2) Moreover, based on the 3-parameter Weibull equation, applying 0.30 MGy HLEBI increased the statistically lowest impact value, as at Pf = 0 significantly 120% over that of the untreated from 29 to 64 kJm−2 indicating increase in reliability and safety.
(3) The improvement in auc could be attributed to HLEBI generating dangling bonds in the low dissociation energy CH2-CH2 and H-CH2 groups (369 and 427 kJmol−1, respectively) of the PP polymer evidenced by electron spin resonance (ESR) spectroscopy measurements showing a peak whose inflection point was at B = ～322.5 mT.
(4) The dangling bonds acted as nano expansion sites from repulsive forces of the outer shell electrons strengthening both the glass fiber and polypropylene matrix along with raising the frictional force at the fiber/matrix interface.
(5) On the other hand, Charpy impact values, auc and the statistically lowest as at Pf = 0 began to drop off when higher HLEBI irradiation levels of 0.43 and 0.65 MGy were applied. Therefore, carefulness was highly recommended to optimize HLEBI dose for maximum reliability and safety when treating CSM GFRTP parts in industry.
(6) HLEBI strengthened the GFRTP surface after molding and shaping with high reproducibly was successfully developed to reinforce the outside of about 100 μm depth of fibers and polymer, as well as fiber/polymer interface adhesion. The 2nd important point to reduce the cost was that the HLEBI-dose of GFRTP surface irradiated was half of that of GFRP bulk by HLEBI-activation of GFs prior to dipping in the PP.
Authors would like to thank Prof. Dr. Akira Tonegawa, Dr. Tadae Morishita along with Hironori Satoh, Ryuichi Suenaga, Noriyoshi Miwa and Naoto Hironaka for their useful help.
