Contents lists available at ScienceDirect Sone i] Composite Structures ELSEVIER journal homepage: www.elsevier.com/locate/eompstruct Residual impact strength of carbon/epoxy laminates after flexural loadings AM. Amaro", P.N.B. Reis”, M.A. Neto®, J.M. Cirne* ‘ceMuc Depart Mechanical Engineering Unversity of Cimbr, Ru ae Rei Sats, Pal de Mracr, 1010768 Coimbra Portal > Depart lectrmecaral Engineering, Overs of Be inter, Cv, Pores ARTICLE INFO ABSTRACT ‘lable ‘Very few studies can be found in the Iferature about the impact strength of laminates containing pre-damages promoted by loads susceptible to occur in-service. Therefore, this work intends to study the impact strength of carbonjepoxy laminates containing pre-damages promoted by flexural loadings. For this purpose laminates were submitted to bending loaés about 35%, 47, 85% and 94% of the ultimate ‘equal Strength To evaluate the presence of the damages, acoustic emission and C-can techniques were used, The magnitude of the intial damage has a significant influence onthe ispact parameter values and ‘on the other hand. the impact strength of laminates subjected to repeated low velocity impact loads is Pedamees also strongly dependent on the damages promoted by the diferent bending loads 1 2006 Elsevier Li. Al 1. Introduction carbonjepoxy laminates containing delaminations, however their ‘Composite materials have been increasingly used but, as conse- ‘quence oftheir poor tolerance to damage, there are some applica- tion problems. In fac, strength, stifiness and global performance of those materials are significantly affected by geometrical and mate- tial defects resulting from an imperfect manufacturing process or from external loads actuating during the operational life (1), In terms of external loads, ow-velocity impact events can occur in-service or during the maintenance activities and they are con- sidered one of the mast dangerous loads for composite materials ‘The impact energy is absorbed by internal damage mechanisms promoting the interaction of several damage types without exte- Flor signs [2.3]. Matrix cracking, fibre fracture and fibre-matrix debonding are typical damages, but delaminations are the most important damages. because they affect dramatically the perfor- mance of composite materials and, simultaneously, they are dificult to detect visually 45]. Delaminations are prone to occur at interfaces between different oriented layers due to mismatch bending that lead to development of interlaminar shear stresses at those interfaces (6) Several studies can be found along the literature about the delaminations’ effect on the residual properties. For example, Reis et al, [7] found reductions of ultimate tensile strength about 16% in 7 caespanaing author -Ematlodéreses ana amaroecem vce (AM. Amaro), pes0upt EN Res supustanetobents pt (MA Neto, jvecmeBers 9° UM. Cie, up bot or20.2015pcompstr2016.03.005, sizes do not influence significantly the value of tensile strength. ‘According to Mosallam ef al. (5) for cross-ply laminated compos- ites, a decrease of about 25% was observed on the ultimate tensile strength after an impact energy of 68 J. This reduction detected on the tensile strength can be explained by the degradation of the fibre/matrix interface and by the stress concentration promoted by the delaminations (9.10). Relatively to the flexural properties, reductions around 34% for [0, 90, 0, 90}, and 78% for [0. 90]. car~ bonjepoxy laminates were found by Amaro et al. 11). Authors explained these differences due to the lower flexural sifiness of the antisymmetric layup relatively to the symmetric one. On the other hand, the delamination position along thickness shovrs to be also important, because the most critical situation corresponds to the delamination located at the mid-thickness of the specimen in consequence of the alterations induced on the shear stress pro- file [12]. Finally, compressive strength is significantly lower than the tensile mode, as consequence of the failure mechanisms [9.1317], but, after impact loading, reductions about 60% can be feached [418.19], According to Suemasu eta. (20), the main cause is related with the existence of multiple delaminations that inter- act during compression, but, for Lee and Park [21], delaminations tend to grow rapidly under post-buckling loads, causing further reductions in structural strength and leading to critical structural failure. According to Amaro et al. (22) the load carrying capacity of composite laminates decreases due to the presence of delamina~ tion independently ofthe layup or type of delamination. The sameauthors concluded that the compressive strength is strongly alfected by the delamination size. It was verified that the number and the size of delaminations plays an important role on the com- pression alter impact behaviour of composite materials, Three types of delaminations, embedded, through-the-width and partial ‘were analysed. They observed that, according tothe type of delam- ination, different buckding modes appear. The local buckling mode, ‘which is caused by embedded and through-the-width delamina- tions, is highly influenced by the layup. The stacking sequence ‘determines the deflection response to the applied load, principally ‘when there are multiple superposed delaminations I the mechanical performance of composites after low-velocity Jnpact loads is conveniently reported along the open literature, few studies can be found about the impact strength in laminates ‘containing pre-damages, in fact, composites can encounter large ‘out-of-plane displacements andjor loadings (such as hydrostatic force) during storage or service, which creates flexural loads on the panels. Therefore, the aim of the present work isto study the impact strength of carbon/epoxy laminates, previously submitted to bending loads. The results are discussed in terms of load-time, load-displacement and energy-time curves. Finally, the damage is evaluated by the impact bending stifiness 2. Materials and experimental procedure Composite laminates were prepared in the laboratory from high strength unidirectional carbon pre-preg (Texipreg HS 160 RM from SEAL*) and processed in agreement with the manufacturer recom- ‘mendations, using the autoclavejvacuum-bag molding process The processing setup consisted of several steps: make the bag ‘and apply 0.05 MPa vacuum:; heat up to 125°C at a 3-S °C/min rate; apply a pressure of OSMPa when a temperature of 120-125 °C is reached: maintaining pressure and temperature for {60 min; cool down to room temperature maintaining pressure and finaly get the part out from the mold, ‘The laminates were manufactured with unidirectional carbon layers with the following stacking sequence [0,903]. The overall dimensions of the plates were 200 x 300 x 1.2 (mm) The quality ‘control of the plates was performed by C-Scan using a Physical ‘Acoustics Corporation (PAC) ULTRAPAC I water immersion ‘C-Scanner, in order to evaluate the eventual presence of defects resulting from manufacturing process. For the ultrasonic analysis a 25 Milz broadband immersion piezocomposite transducer was lused in the pulse-echo mode. The same transducer received the ‘echoes originating from multiple reflections inside the specimens from a longitudinal normal incident wave. The green colo, present in all plates, is related to the backwall echoes, evidencing the absence of any damages [23] The specimens used in the experiments were cut from these thin plates, using a diamond saw and a moving speed chosen to feduce the heat in the specimen, with dimensions of 100 x 100 1.2(mm) and submitted to flexural tests. Three- point bending (3PH) tests were carried out and the procedure ‘was adopted according to ASTM 790-10, using a Zwick universal testing machine, model 1435, equipped with a 5 KN load cell ata displacement rate of 3 mmjmin and a span of 50 mm, The flexural tests were cartied out on similar specimens to those used in the impact tests, because the damage was correlated directly with ‘the load, After obtained the ultimate flexural strength, some speci fic Toads were selected and the specimens tested with them, in ‘order to promote pre-damage. For this purpose, the flexural tests ‘were monitored with the AE technique. Four loads associated to the low amplitude, middle amplitude and high amplitude events, were selected. After that, all specimens were evaluated again by (C-Sean system described previously. A Marandy acoustic emission analyser, model MR1004, was used to provide the amplitude and the number of ringdown counts for each acoustic emission (AE) event. This equipment has a threshold voltage adjustable from 10 mV to 69.18 mv. The AF event amplitude is defined as the maximum [AE signal level relative to 10 mV, The amplitude detector unit sorts ‘the AE events into twenty-five levels, each one with a bandwidth of, 2448, Level zero corresponds with the amplitude above 10 mV and below 13.18 mV. For ringdown counting, the amplitude of ‘the signal is compared with the threshold voltage. and the total ringdown count for each event is the number of times ofthe signal amplitude exceeds the threshold level. More details about this equipment can be found at (24) Finally, multi-impacts tests at low velocity were performed using a drop weight-testing machine Instron-Ceast 9340 and a 10mm impactor diameter with a mass of 3.4 kg was used. The tests were performed on circular samples of 70 mm diameter and the impactor stroke at the center ofthe samples obtained by cen- trally supporting the 100 x 100mm specimens. For all tests, an impact energy of 1 was used, which corresponds to an impact velocity of 0.77 ms"*, For each condition, three specimens were tested at room temperature. 53, Results and discussion Flexural strength was studied and ig. 1 shows the load= displacement curves obtained. An offset of 0.2 mm was used for clarity. The plot shows a nearly fragile behaviour, with a line region up to the maximum load (Pra). and 2 significant drop of the load after Prax Was absetved. The mechanism of damage agrees with previous works developed by Amato et al. [11] and Reis etal [25), where the fracture of the fibres, in compression, with quite small delaminations around the broken fibres can be observed. ‘The zigzag aspect ofthe curves showed in Fi. | results from fibres breakage. The posterior load drop isa consequence of the propaga tion of delaminations initiated at the regions with broken fibres, According to Reis etal. 25], the high compressive tress concentta- tion in the pin load contact region associated with the low compressive strength ofthe fibres promotes compressive breakage of the longitudinal fibres in this region. In terms of ultimate flexural load, an average value around 1923 N was obtained with a standard deviation of 28 N, In order to promote the pre-damages in the laminates, and according with the acoustic emission events obtained ftom the Mlexural tests, four loads were selected: 700N (associated to the low amplitude events); 900 and 1250.N (associated to the middle amplitude events); and 1800.N (associated to the high amplitude events) 1200 00 400 ig. 1. Load displacement curvesAM Aare ea Compose Sracues 145 (2018) 69-74 ” ‘These values are, respectively, 36%, 47%, 65% and 94% ofthe ulti- mate flexural strength. Fig. 2 represents the number of events ‘obtained for each sound intensity and load level. These values were ‘obtained for a pre-load of 1800 N, however, they are representative of al tests performed (independently of the pre-load level). Tris possible to observe that the AE activity starts atthe begin- ning of he loading history, indicating the early occurrence of dam- age, Therefore, the oad of 700 N was selected in order to study the damages associated to the low amplitude events, as shown in Fig. 2. Low amplitude events are the first events that appear and, according to the literature |24.26-26), they are associated to the ‘matrix plastic deformation and cracking, crack propagation and ‘Sbre/matrix interfacial debonding. The propagation of the matrix cracks contribute for the debonding fibre] matrix resulting mainly from triaxial stress state developed inside the composite [24], Rel- atively fo the damages associated to the middle amplitude events, twa loads were selected, One of them for a low number of events (900 1) and other one for a high number of events (1200 N) corre- sponding. in this case, to damages with different severities, Delam- inations are responsible by miidle amplitude events (28, and they ae related to the presence of matrix cracks, which promote stress Concentrations and induce delaminations atthe interface between the cracked ply and the neighbouring ply [2-22 Finally, the load of 1800 N was selected because itis associated to the high ampli- tude events, which are the result of fibre breakage [2425-28] Fig. 3. shows the damages obtained by C-Scan. From the pictures, tis evident that until 800 N (inclusively) only the green colour occurs and, consequently, this technique is not able to gx caine 1 5% seo, 5 00 wo 1600 —2ho Load] ws ay m0 | alice 40 se 5 ° 02M) Lod detect the damages. Infact, the damages reported in Fig. 2 for these load levels are difficult to detect by standard C-Scan techniques, because they do not offer a wide enough reflecting surface a3 delaminations (33). However, after 1250N (inclusively), the damage can be detected without ambiguity and different colours are clearly visible to identify the defect, In terms of Fis 2, corresponding to the load of 1250, itis possible to identify two colours, while the Fig. 2e) presents three regions with diferent colours. in this case, the blue region represents the near-surface defects and the red area corresponds to the middle depth discontinuities. After the damage induced by the flexural loads mentioned above, the samples were submitted to impact loads, and the results ‘compared with those obtained for the control samples (ON). Fig. 4 shows typical load-time and energy-time curves obtained, In terms of load versus time curves, itis possible to observe that the load increases up to a maximum value (Pax) followed by a drop after the peak load, In all tests the impactor deforms the spec imens and always rebound, which means that the maximum impact energy was not high enough to produce full penetration [34-37], From the curves that represent the evolution of the energy with time, the beginning of the plateau of the curve coincides with the loss of contact between the striker and the specimen, so this energy coincides with that absorbed by the spec imen [25-38]. The elastic energy (rebounded energy) can be calculated as the difference between the absorbed energy and the energy at peak load (25.36), and itis defined as the energy stored elastically in the specimen and transferred back to the impactor 25) Table + presents the average values, and respective standard deviation, relatively to the maximum load, maximum displace- ‘ment and elastic energy (rebounded energy). As expected, these parameters are significantly affected by the magnitude of bending loads. Higher pre-loads are responsible by higher damages (as shown previously) and, consequently, the strength and stiffness ofthe laminates decreases, Therefore, this explains the decreasing of the maximum load and the increasing of the maximum displace- ‘ment observed, These results agree with Amaro et al. [40], where a strong dependence of these two parameters with the stiffness was found, n terms of elastic energy, the increasing of bending loads promotes lower rebouncled energy, and this tendency is similar tothat observe forthe maximum load, The damage accurmulation that occurs, tween the pre-damages and the damages induced by the impact loads. is responsible by the increasing of the absorbed energy in the specimens, and consequent lower elastic energy observed [5.29] ‘Compared with the control samples, when the specimens were submitted to bending loads of 700 N, the maximum load decreases about 1.1% and the elastic energy around 2%, but the maximum displacement increases 12%. However, these values increase to 25%, 48% and 24%, respectively, when they were submitted to bending loads of 900 N, These results confirm that, despite the fact that C-Scan doesn't present detects for these levels of load, the specimens contain small damage as shown by the AF technique, However, when the damages are detected by the C-Scan, the resid ual impact strength is significantly affected, For example, the max imum load decreases around 28.1%, the elastic energy decreases around 27%, but the maximum displacement increases about 52% when the specimens were submitted to loads of 1800. These results are in good agreement with those reported by Amaro etal, [41-42) in previous works ‘On the other hand, the impact bending stiffness is an important property to assess the damage resistance of a composite laminate, and it can be obtained by the slope of the ascending section of the load-displacement plot. However, more details to determine this property can be found in Ref, [43]. Table 2 presents the impactAM Amar tl CompasteSacres 14 (2015) 68-74 @ ) © @ © "ig 3, Typical inveral damages observed by the Scan ehniqu for specimens submited to Bending loa of 2] ON (8) TON: f) 900 (1250; 1800, 02 4 6 8 WD 1s 6 I 20 2 Time rs] toe wd ere Uns etna ater being bending stiffness, measured after the first impact. This parameter shows to be highly dependent of the bending loads and the tendency observed agrees with Figs. 2 and 3. For example, rela- tively to the control samples. the impact bending stiffness decreases about 23% when the laminates were subjected to bending loads of 700.N, but this value increases to 46.5% when the laminates were submitted to bending loads of 1800N. “The effect of multi impacts was also studied and Fig. 5 presents, for the different bending loads, the number of impacts to failure. ‘Mean curve, fitted to the experimental results by least squares Experimental resus obtained rm the pct et method, is also superimposed. The laminates were considered failed, when full perforation (FP) occurs, Full perforation is defined ‘when the impactor completely moves through the samples. It is possible to observe that the resistance of the laminates to repeated low velocity impacts is very dependent of the bending loads, In terms of average number of impacts to failure, and comparatively to the control samples the damage promoted by the bending loads ‘of 700, 900, 1250 and 1800 induces a decreasing about 19.5%, 299%, 48.1% and 71.4%, respectively The damage promoted by the bending loads decreases the clastic energy, and a similar effect can be observed when the lam- inates are subjected to multi impacts, as consequence of damage ‘accumulation [41]. According to Amaro et al. (44), this damage ‘accumulation can be expressed through the impact bending stiffness. and this property is plotted in Fig. 6 against the impact number. The curves that show the absorbed energy against impact, number are also superimposed, for control samples and for specimens subjected to bending loads of 900, 1250 and 1800. ‘The last impact is not presented because the full penetration ‘occurs, These curves are representative of all conditions tested, ‘Notwithstanding the opposite tendency observed. in both cases, different stages can be identified as a consequence of damage accumulation (44), Relatively to the control samples and samples subjected to bending load of 900N, three stages can be found. For example, in terms of control samples, the impact bending stifiness decreases around 73% during the frst 40% of the total ite, followed by a slow decreasing (about 6%) during the second stage Finally, in the thied stage, the impact bending stiffness drops ‘Vales of tbe being ons on 7008 Soon TON Ta20N ‘Maxima tov) ony 76) 75 25) waa 380 Maxima esplacement asa] Bsa) 2s 31G3) 3s (a2) 3803) asc enetzv 31 50135) Saocan) srGe, 5a 39) 3655) {TSaneard cevation inAM Aare ea Compose Sraes 145 (2016) 69-74 2 inst wih es a a +0 me ON 4 900 508 mE IRIN < oe er oe ot cracking isthe st damage, which soloed by 2 rapid dani 0 Savon propagation On the othe: hand. waen he innate were suerte to bending loads higher than 1250 nee) sage ° oe Ts subsite bythe stage Ia consequence of he damages every prevents The stp in ths case ate no aes Lncnrnenennnencennen! tytn induce the micocrck esponsble bythe damnation, Fig 5 Elec of he being loads onthe impact sengh Name of pact [N] suddenly. tn terms of samples subjected to bending load of 900 N. a similar behaviour can be found, but the first and second stages are shorter and more abrupt. Three distinct zones were also found by AAzouaoui eta. (45), which are characterized by initiation and mul- tiplication of delaminations (zone 1), saturation of delamination (one Ml) and ply cracking with fibre breaking (zone If), According tothe authors 45), in zone I the delamination appears inthe frst cycles of shock loading, which occur initially in the interface ofthe last layer. followed by new delaminated surfaces in other inter- faces. After a certain number of impacts. the phenomenon of delamination saturation appears in which the propagation of delaminations and the development of new ones are stopped. Finally, zone Il is characterized by an acceleration of the damage until final falure and this stage of damage is characterised by the failure of fibres. However, delaminations are consequence of several damage mechanisms, where matrix cracking is ehronolog- ically the first damaging mode. Studies developed by Schrauwen and Peijs [46] and Boukhili et al. (47) prove that the matrix but they promote the rapid delamination propagation. Finally, Fig. 7 shows the repeated impacts effect on the maxi- mum load (Prue) and maximum displacement, where N is the number of impacts at any given instant of the test and Ny is the ‘numberof impacts to failure. The dimensionless format was used for clarity. It is possible to observe that the maximum load tends to decrease, following the same tendency observed previously. For bending loads lower than 900N (inclusively) there are three stages, while for higher values, there are only two stages. Once ‘more, the number of stages is related with the damage severity promoted by the bending loads. Therefore, the damage accumula~ tion responsible by the lowest stifess of the laminates can explain the decreasing of the maximum load, However, for high impact numbers, the maximum load drops dramatically because the fibre breakage decreases the local rigidity at the point of impact [46]. In terms of maximum displacement, Fig. 7 shows an opposite Kendency, where the damage accumulation explains, again, the increase observed. 4. Conclusions ‘The impact strength of carbonjepoxy laminates previously sub- titted to bending loads was stidied, Simultaneously, the acoustic emission and the C-Scan techniques were used to evaluate the damages promoted by the different levels of bending loads. It was possible to conclude that the last technique was not efficent to devect the damages promoted by the lowest bending loads, but, for the highest loads, both techniques show to be able to detect the damages. However, they are complementary to under- stand the diferent damage mechanisms that occur along the experimental tess. The magnitude ofthe damages promoted by bending loads has a strong influence on the impact strength, For example, the inereas- ing of the bending loads induces a decreasing ofthe maximum load, but an opposite tendency occurs in terms of maximum displace” ‘ment and absorbed energy. In terms of multi impacts, the number of impacts to failure is very dependent of the bending load values and the damaged area can be correlated with the impact bending Stifiness, Simultaneously, the multi impacts promote a decreasing ofthe elastic energy as consequence of damage accumulation.” AM Amar tl Compas Sacres 14 (2016) 68-74 Acknowledgments ‘This research is sponsored by UID{EMS/00285)2013. References In] Rei PNB, Fee JAM Antunes FV, Costa OM, ll crack length on the ‘Recanar actte ol moves cabeniepny abies Res en [a] Adams RD. Casey FD. A review of detects types and nondestrucve esting ‘edhe fo empostes ane boned Jts MDT int 198:21 208-22, Ip] Amato ane Hels PMH de Moura MISE. Santos |B. Damage eeeccon on laminated composte maceals using several NOT techniques sgh la] Avrae 5 impact on laminate composite materials Appl Mech Rev Iorssn sooo, Is] Reraran SOW, Wiseart Ne Review of low-velocty impact proper of emposte mates. Compes Part A Apl Sc Manu 198.27 113-31 Tol de Nurs MISt, Conaver Jt Moueing he Interaction beeen matrix {Caching and delamination In eatoncepony laminates under Taw Yl) Smee Compor se Technol 20bt 1057 In] Roe PNB Foci J, Antunes Vy, Riharvon MOW. Hct of neayer| ‘etna nme Sear xbegey its, J Camps [a] Most A Sen J Wine J Assessment of residual tense sength of elena compnites secede Towers ac | Rea Bi Ip] Mite Roman Tee popes of real uniecional kevaepaxy feurpostes Composites 1990.21 63-2 tio) Otiten Te Agunout Ne Zana € Tension aigue analysand te oedcion for compost anintes it asgue 1999 11379-33 Int] Amaro AM de Moura ESE. Res PN. Residual stength afer low velocity lnpace in carbobepeny lamaates. Mater Se Farun 2006/514-516 820-8. [12] vara AMC Res PNB. ce Moura MSF. Delamination ete on bevding Benaiour la eatboo-epory composts sian 201147 205-8, Ina] Notansht¥, Hapa i amaca HFaconaphy of facture in CFRP under Compressive loa Compos se Techel 1987°97113%87. Ina] Laatore) shear versus elataconal amaze rwechanisms in che compcesive fare of fbreenforced composes Compes Part A\ A Sel Manuf Sosrowaisony, Ins] bering F. Sots C. Gur IA. Comprecive fire of 0 res uniirecional ebentbereintered plastic (CPR?) Taminates by Aire merobuesing Campos Se Teco! 1998591455. Ing] Lee it Yeramal cS. Wass AM, Compressive spiting respanse of liber [17] kes NB, Ferrer JAM, Costa JOM Richardson MOM Fatigue ie evaluation fer carbonfepony laminate composites under constant ahd vanisle block louding Campos Se Teco! 20085815450 Ui) Bienes SiN Blanton actin aad sete Eg Fat Ins] de Moura MES. Marques AT. Predict of ow veld impact damage ln {lsben-epony lalate. Cones Pat A Appl Se Maa 200233 381-8" [ao] Sess Sass Wishaw Z, Aad ¥-A ume st 98 compesive Behaviour of composte plates with ‘hulle cheat deaminations| Considering delalaatin propagation Comps Technol 2008 68 2362-7, [any Lees, eri D-¥. ucklng nats of laiate composes lates containing felanunatins using ennacedasuea sai sol element int Sos tact Sara, AN, eis PNB, e Moura MPSE Net MA, Bucking analysis of aunated ste plates submited to compression ater impact. Pers Pom s6he8s [ANE Ree FN, Magalies AG, de Moura MESE. The Inuence ofthe Boundary cangiion? on low-veloety impact compente mage. Stain bags a20-8 [aa] esr ec JANefiharésn MOW. Fatigue damazecaractrizaien by NOT in patptopyienejless nce composes App Compes Mater 2011:18 ote 125] Kes FN, Fecera JAM Antunes, Cost JON lex behaviour of yb Temnated composes Compes Fae A Appl Se: Mana 200738 1812-20, [26] ée creat Fl. Winen PAM. fnssen RSP. eal eqoeney ceteriaton of Eotstie essen for aieent acure mechanisms exDOn/epeNy empenter Compor Se Teeel 1895'5:405-12, lar] Hague 8 Godin Garner Salman Villar Ure of acount misono sd age nates nh ested ee: Cpa Sa Teed 128) Zhuang X Yan Xlavestigation of damage mechanisms in self-seinorced Davehylene composes by acoustic estan. Comper Si Teche! 200665, Greece [29 jets Chang F.Casactszton of mati eck induced false ‘ipementjCompes Mate 2001352005-35 (30) ftsen Phang Characters of rai cack induced fale — part {uly snd version | Compr Mate 2001 252037—74. [st] Hale si ang whan B, Worn MI Medel te stration Between Inatoe cack apd delaminacen damage in scaled quastisouope spears {Compes Set Teche! 200858 80-9. [pz] Zuniga Turon A, Mato. costa Mabel. unde. An ence base are ‘ten fr mare crac induced elanaton in laminae compost Structres:Compos se echne 2014:112°338-84 [a3] Asmench Felt Utaronic elation of mats damage in impacted feimpesite laminates Conpes Part 8 Eng 2000316, Iba) Ketan Karan Obtain 8 The responce af laminated compose plates tee oweveoty pact lacing. Cortes suc 2003:58 118-77, [bs] kesePNB Ferre JAM Sasa Hcharcon MOU Sonor npatesponse ff Kev: composes with led epoxy matic Compos Sibet 2012:94 Ssaa-8" [36] Res Pi, Frees JAM, Zhang ZY, BenameueT,Rihedron MOM. Impact, spi evar pens ith ancy ened poy mat Caos [a] Res ING Fein AML ag 2, eT, Rion MOW np [ss] Rfo To Zaera& SuibetoE Navete C Damage n CFRPS duet ow velocity Input at lo temperate compos Part 8 Eng 2005 3641-80 [39] King tan Cer absorp mechani is Kevlar malta wap fic composes andes impact leading Comps Se Teh 200:50773-44 [ao] Aine AME Res PMH de Maus MSE Sane nflunce ofthe specimen thikness on Tow veloey unpact Behavior of composts.) Polyn ED [at] Aiace AN Rel PNB, de Moura MEE. Neto MA. tence of memnpact on (ae composts laminates Compes Pate 8 Eng 2013 2°93-5 laa} Aoaco AM, he PRB. ce owes MISE. Neto MA Inflsence of open holes an Composites delamination induce bylaw velocity inact lds. Compas SUM Sores aise [4s] Diswet OS Nash OM, Banks WME An experimental sty of damase ccumslasen imbalanced CP? aminates deepened pac Comes Stor 2006 832475, [a4] Amara AN dir PN, Neto MA Laure Cele of alain and ci slaione fr aasieponyeompestes. Polym Degra Stab 20139885352, Us) Reinke» Aa fBemncatene ati A, Png trae ftgu Inj Fatigue 200122877 45 146) Scauwen Pee infor nati act nd ce architecture onthe Tepes inpedt respons of gsefhe-enicend lasted compestes "opi Campos Mater 20028 331-82 [a7] Boast Bt C. causia Fuge mechanisns unde ow eng epeated ‘pat of cmposte laine j Rein Past Compos 1998 13.656-70,