Multi-scale analysis of the generated damage when machining pockets of 3D woven composite for repair applications using abrasive water jet process: Contamination analysis

Texte intégral

(1)Multi-scale analysis of the generated damage when machining pockets of 3D woven composite for repair applications using abrasive water jet process: Contamination analysis Xavier Sourd, Redouane Zitoune, Akshay Hejjaji, Mehdi Salem, Laurent Crouzeix, D. Lamouche. To cite this version: Xavier Sourd, Redouane Zitoune, Akshay Hejjaji, Mehdi Salem, Laurent Crouzeix, et al.. Multiscale analysis of the generated damage when machining pockets of 3D woven composite for repair applications using abrasive water jet process: Contamination analysis. Composites Part A: Applied Science and Manufacturing, Elsevier, 2020, 139, pp.1-11/106118. �10.1016/j.compositesa.2020.106118�. �hal-02941045�. HAL Id: hal-02941045 https://hal-mines-albi.archives-ouvertes.fr/hal-02941045 Submitted on 7 May 2021. HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés..

(2) Multi-scale analysis of the generated damage when machining pockets of 3D woven composite for repair applications using abrasive water jet process: Contamination analysis X. Sourd a, b, *, R. Zitoune a, *, A. Hejjaji a, M. Salem c, L. Crouzeix a, D. Lamouche b a b c. Institut Clément Ader, CNRS UMR 5312, 3 Rue Caroline Aigle, 31400 Toulouse, France Safran Aircraft Engines, Villaroche, Rond-point René Ravaud, 77550 Moissy-Cramayel, France Institut Clément Ader, CNRS UMR 5312, Campus Jarlard, 81013 Albi, France. A B S T R A C T Keywords: A. Polymer-matrix composites (PMCs) A. 3-Dimensional reinforcement D. Surface analysis E. Machining. Repair of damaged aircraft structural parts is a regular maintenance activity carried out to preserve airworthiness while being economically viable. Conventional repair techniques used have negative impact on the repair quality and environment due to dust emission. For eliminating aforementioned problem, this study investigates controlled depth abrasive water jet (AWJ) milling for repair procedure of 3D woven CFRP composite structures used in the new aircraft engine developed by Safran Aircraft Engines. The aim is to identify and quantify the damage and surface contamination induced by AWJ milling. For this, the influence of milling parameters on the damage and surface contamination is investigated through full factorial experimental study. Several character isation techniques like 3D optical profilometry, Scanning Electron Microscopy and X-ray tomography combined with image processing were used for a multi-scale analysis of the machined surface damage and contamination. In fact, the machining quality was quantified using an innovative criterion called “crater volume” (Cv). The obtained results have shown surface contamination by embedded abrasive particles and presence of bare and broken fibres, cracks, craters. The results show that the rate of contamination is influenced by the jet pressure and the scan step and the Cv is influenced by the jet exposure time.. 1. Introduction Due to the more and more stringent environmental rules provided by the aviation authorities, the aerospace industry has to drastically reduce the fuel consumption of the aircrafts. In order to achieve this goal, manufacturers decided to introduce advanced materials which have a high strength to weight ratio in the aircraft structure. To do so, they have developed cutting edge composite materials composed of polymer ma trix and 3D woven fibres, used either for monolithic or multi-material structures. In service, these structures can be damaged, mainly by erosion or impact concerning outer parts, and are rather repaired than replaced during maintenance due to the high material and manufacturing prices. However, the repair procedure is subjected to strict guidelines provided by the maintenance and repair organizations (MRO), making it expensive for the aviation companies (aircraft grounding, highly skilled staff). In addition, as the composite repair are more and more performed by adhesive bonding, the quality of the. machined surface is of primary concern to have a strong and enduring bonded repair. However, the quality of the machined surface is strongly influenced by the process of machining as well as the machining pa rameters [1–3]. In the industrial field, conventional machining is used for material removal to remove the damaged area. Nevertheless, Carbon Fibre Reinforced Plastics (FRPs) are difficult to machine properly because of their heterogeneous mechanical and thermal properties. In addition, due to the high abrasiveness of carbon fibres, important and premature tool wear is recorded during conventional machining of CFRPs, leading to high machining temperature and hence thermal degradation of the matrix, which has a low thermal conductivity compared to carbon fibres [4–6]. Another important drawback when machining CFRP by conventional means is the emission of harmful particles in the air, in form of carbon fibre dust, which is dangerous both for the environment and the operator’s health [7,8]. Moreover, the main mechanical defects produced by conventional machining of CFRP are in form of delamination, broken fibres, fibre pull-outs and craters. * Corresponding authors. E-mail addresses: xavier.sourd@iut-tlse3.fr (X. Sourd), redouane.zitoune@iut-tlse3.fr (R. Zitoune)..

(3) [4,5,9,10]. The location and expanse of all these defects mainly depend on the process parameters and the machining direction with respect to fibres orientation [11,12]. Both the thermal and mechanical defects can lead to a drop of the mechanical properties (endurance limit) of the structures in service [6,7,13]. Based on all these issues, the use of non-conventional machining processes has been considered for CFRPs, such as laser beam, electrodischarge or chemical machining. By removing the rigid tool in the process, cutting forces are considerably reduced, making the workpiece more easily positioned on the machining table. However, laser beam machining generates a high temperature, resulting in the presence of Heat Affected Zones (cracks, burns) and fumes which can be harmful for the operator if inhaled [14–16]. This process has also a low material removal rate compared to other techniques, increasing the machining time which is not suitable for economic reasons [17]. Electro-discharge machining produces also a high temperature, leading to large heat affected areas, and other defects as fibre swelling, delamination and recast material [18,19]. Moreover, this machining technique greatly depends on the material’s conductivity, making more efficient for metallic materials applications. Chemical machining, currently employed for metal-composite structure machining applications, is a time consuming and dangerous process both for the operator’s health and the environment because of the corrosive chemical products used. In order to avoid these issues, abrasive water jet (AWJ) machining has been suggested as a substitute non-conventional process. The water jet lowers the spread of carbon particles, by clearing out the chips and dust, and eradicates the heat affected zones [17,20]. Moreover, this machining process permits to remove a constant depth even on slightly curved parts [21,22], operation which can be complicated when machining is per formed by conventional techniques. Numerous studies [17,22–25] have shown the versatility of AWJ which can machine a lot of different ma terials, such as titanium alloys and FRP laminates composites. To the authors knowledge, no study has been performed concerning AWJ machining on 3D woven composite, all the less the contamination phenomena of composites (unidirectional laminates, 2D and 3D woven fabrics) by abrasive grit. In addition, the mechanisms of material removal are independent on the fibres orientation because they are performed by micro-erosion consecutive to impact of the abrasive par ticles [26]. AWJ, by overcoming many drawbacks of conventional milling, seems then to be an interesting alternative process, especially in case of composite repair application in the aerospace field. However, as for all machining techniques, AWJ generates several kinds of defects. On one hand, it has been shown by Hejjaji et al. [22,25] that milling with this technique on CFRP laminates made of unidirectional plies induces defects and damage in form of broken fibres, fibre pull-out, fibre-matrix debonding, craters, abrasive embedment and grooves, which location and expanse depend on the choice of the machining parameters. All these defects are due to the material removal occurring mainly by microerosion of the matrix and brittle fracture of the fibres consecutive to solid impact of highly energetic abrasive particles [3,22,26–30]. On the other hand, the energy carried by water can favours delamination phenomenon by wedge effect if the parameters are not well chosen, which makes plain water jet machining unsuitable for composite mill ing. In case of AWJ, the abrasive flow rate must be carefully selected so the fraction of energy transferred from the water to the abrasive is important enough [21,24,31,32]. Besides, the high speed abrasive par ticles produce another kind of defect when impinging the machined surface. Once embedded in the target material, they introduce stress concentration zones which weaken the structure, by getting inside the cracks and avoiding them to close for example. As for the other types of defects, the degree grit embedment depends on the machining param eters – mainly on the energy transferred to the particles – and the milling path as well as the nature of the material [33,34]. Indeed, the machining strategy has to be properly designed in order to avoid sharp direction changes which can modify the erosion conditions and then worsen the surface quality. It is important to notice that the classical criteria used in. industry to quantify the post-machining surface quality are the average surface roughness Ra and the surface waviness Wa. In case of AWJ milling of CFRP laminates, Hejjaji et al. [22,25] have shown that Ra depends on jet pressure and traverse speed in order of significance, with low influence of scan step. Likewise, the surface waviness of the milled specimens depends on scan step, jet pressure and traverse speed. How ever, surface roughness does not seem to be an efficient criterion to link surface quality and post-machining (conventional or non-conventional techniques) mechanical behaviour of CFRPs, both in static and fatigue, because of contradictory results [35–37]. In order to overcome this issue, another parameter has been considered, called “crater volume per unit area” (Cv). Contrarily to the classical parameter (Ra), Cv permits to take into account and quantify craters, the main defect encountered when machining composite with AWJ. Moreover, it has been proven than as the crater volume increases, the mechanical properties of the machined specimens are altered [22,35]. However, no work, to the authors’ knowledge, has been performed concerning defects identifica tion and quantification consecutive to AWJ milling on 3D woven com posite specimens. The goal of this work is then to identify and quantify the damage induced by abrasive water jet milling on 3D woven CFRP. The influence of the machining parameters of the process on the different nature and sizes of damage will be studied. To do so, specimens made of 3D woven CFRP have been milled following a full factorial experimental design of three machining parameters (viz. water jet pressure, traverse speed and scan step). After the milling phase and based on several characterisation techniques like 3D optical profilometry, Scanning Electron Microscopy (SEM) and X-ray tomography – a multi-scale analysis of the machineinduced damage with respect to the process parameters is discussed. A focus is made on grit embedment and the effect of the process param eters on the contaminated area. Finally, the “crater volume” parameter (Cv) expressed in terms of volume per unit area is used to quantify the machined surface defects. 2. Material and methods 2.1. Material The tests of this study have been performed from a plate made of CFRP 3D woven composite. The epoxy resin and fibres constitutive of the material are respectively referenced as PR520 (Cytec Company) andIM7 (Hexcel Composites Company). The plate was manufactured by light Resin Transfer Molding process, having a thickness of 9.75 ± 0.04 mm and a nominal fibre volume fraction of approximately 54%. No further details (e.g. weaving architecture and mechanical behaviour) are presented due to the confidentiality of the material, which is the prop erty of Safran Aircraft Engines. As seen from Fig. 1, the surface of the plate is rather flat (±30 µm from the 3D topography) with minor scratches due to handling and transportation. A full factorial design has been selected, with 36 specimens (140 mm × 20 mm), all originating from the same plate and obtained by abrasive water jet (AWJ) cutting. For each specimen, a given set of parameters (viz. pressure, traverse speed and scan step) was selected to mill by AWJ three identical pockets, of size 20 mm × 20 mm. 2.2. Abrasive water jet milling machine and parameters All the machining operations (trimming and milling) were per formed on an AWJ machine Mach 4c from Flow company. The different parameters (fixed and variable) selected for the tests are gathered in Table 1. Some parameters (such as the cutting head geometry) are fixed for economic reasons. The abrasive employed from the machining op erations is a 120 mesh (around 125 µm in diameter) Arabian garnet produced by Garnet Arabia Company Ltd from Saudi Arabia. A single size of abrasive grit was kept for all the machining process due to limitation of the AWJ machine (only one abrasive feeding system available)..

(4)

(5) * 2

(6) 3

(7) /$1 2/

(8)

(9)

(10)

(11) 9

(12) " .

(13) 3

(14)

(15) /*

(16)

(17) /* . J . 1 . J . 1 . ( . /

(18) . /

(19)

(20) .)) + . ; C;> . 1 1 21 3. G@B;;?B;?B ;F>. + . J 2B

(21) 3. DB=B@. !

(22) ! 23. C FB; CB; F. & ). & 8 . !

(23)

(24) . ?> C CD I

(25)

(26) K;DC C ;@ B

(27) ;CC . &

(28) ´

(29) :@<"

(30) . 8 "

(31)

(32)

(33)

(34) )) " *)

(35) + )

(36)

(37)

(38)

(39)

(40) 4

(41)

(42)

(43) ))

(44)

(45) " 8 9* C ;@ B

(46) 4

(47) ’

(48)

(49) 5 .

(50)

(51) 9

(52)

(53) 8

(54)

(55)

(56) . 2!3 :DD"DF<" ;CC " :G"=C<

(57)

(58) &67 + .

(59)

(60)

(61)

(62)

(63)

(64)

(65)

(66)

(67)

(68) &67 ,

(69) . " "

(70) ’ .

(71)

(72) .

(73)

(74) 8

(75)

(76) ’ :DD"D>"=;"=D< 4 . .

(77)

(78) .

(79) 2 + ;3 +

(80)

(81)

(82) 8

(83)

(84) 5

(85)

(86) :DD"DF<" .

(87) 2 + ;3 ,

(88) .

(89) .

(90)

(91)

(92)

(93)

(94)

(95) . .

(96) ’

(97)

(98)

(99) +

(100)

(101)

(102) *

(103) ))

(104) .

(105)

(106) ;? 1 ,

(107)

(108) " . .

(109)

(110)

(111)

(112)

(113) “ ” 0 "

(114)

(115)

(116)

(117) ;CC " . * *

(118)

(119) 9

(120) " 9

(121)

(122)

(123) ))

(124)

(125)

(126)

(127) .

(128)

(129)

(130)

(131)

(132)

(133) ' " .

(134)

(135)

(136) :=< ,

(137) . " .

(138) 9 .

(139) )

(140) + "

(141)

(142)

(143) "

(144) .

(145) ; F ;CC

(146) )). *

(147)

(148)

(149) ,

(150)

(151) "

(152) . " :=="=F<" I

(153)

(154)

(155) . + .

(156) “ ”

(157) .

(158) )) " ; C;> " .

(159) " 9

(160)

(161)

(162)

(163)

(164) + ; /

(165)

(166) " >

(167) " . " .

(168)

(169) "

(170) ;C@ 4

(171)

(172) "

(173).

(174)

(175)

(176)

(177) 2 / D 3 &

(178)

(179)

(180)

(181)

(182)

(183)

(184)

(185) ’ +

(186) .

(187) "

(188)

(189)

(190)

(191) 2 / D3 6

(192) >C " . ) .

(193) 00 $ 5 $ 000 &&$ 6 +

(194) 9 ,

(195) 9

(196) / !(

(197) 2 / 3 +

(198)

(199) 5

(200) " 5

(201) .

(202) *

(203)

(204)

(205)

(206) ’ * + D .

(207)

(208) + .

(209) 5

(210)

(211) 2C ; μ3

(212) 2

(213) ;F

(214)

(215) 3 000 ! 6 ) ! +

(216)

(217) 5 5

(218) 9

(219)

(220)

(221) .

(222)

(223) 5

(224)

(225) . +

(226)

(227) .

(228)

(229) :=< +

(230) "

(231). .

(232)

(233)

(234)

(235) 5 9 “ ” ..

(236)

(237) .

(238)

(239)

(240)

(241) 2G ?F 3

(242)

(243)

(244)

(245)

(246) ,

(247)

(248) "

(249)

(250) 5

(251)

(252)

(253) &

(254) 9 +

(255) 5

(256) '

(257)

(258)

(259)

(260)

(261)

(262) :DD"DF<

(263)

(264)

(265)

(266) 5 +.

(267) .

(268)

(269) 9 9 +

(270)

(271) ;= × ;=

(272) ’

(273) . .

(274)

(275)

(276) . *

(277) +

(278) .

(279)

(280)

(281)

(282)

(283) .

(284)

(285) +

(286) .

(287)

(288) 2 3 !

(289)

(290)

(291)

(292) 23 !

(293) 2/

(294)

(295)

(296)

(297) 9.

(298) "

(299) 3.

(300) 0 9 5

(301) 2/

(302)

(303)

(304)

(305) 9

(306) " .

(307) 3

(308)

(309) 1 9

(310) 5

(311) 1 . J . 0 . J

(312) 2μ3 (

(313) 2μ3 !

(314)

(315) 2D3 L 2D3. ;C* C ; D C ;> @ × ;> @ ;= C × ;= C . !-

(316) 9

(317)

(318) 5

(319) 9

(320) . "

(321) . 5 2

(322) $

(323)

(324) 6 3 .

(325) .

(326)

(327) -

(328).

(329) / =

(330) ;C 9 = μ + .

(331) .

(332) "

(333)

(334)

(335) 2(; (3

(336)

(337)

(338)

(339) 2+; +3 & I

(340) 9 "

(341) D F " .

(342)

(343) .

(344) +

(345)

(346)

(347)

(348) 2$ (

(349) . 6 (3

(350) 22$ +

(351) 6 +3

(352)

(353) 2 3 $ ( = ; ?? ± C D? μ" 6 ( = ?G ± C ? μ" $ + = ; =F ± C ;? μ

(354) 6 + = >F ± C >G μ !

(355)

(356)

(357)

(358) 2!-3 9

(359)

(360)

(361)

(362)

(363)

(364)

(365)

(366) 2 / F3 !-

(367)

(368)

(369) 2!-3

(370)

(371)

(372) 9

(373)

(374)

(375)

(376) "

(377)

(378) "

(379) 24!-3

(380)

(381)

(382) ,

(383) . " 4!-

(384)

(385)

(386)

(387)

(388)

(389)

(390)

(391)

(392)

(393)

(394)

(395) 4

(396)

(397) 4!- 2, 7 3"

(398)

(399) .

(400) 5

(401) 9

(402) .

(403)

(404)

(405) !- ) F × F

(406) .

(407)

(408)

(409)

(410) .

(411) + - + ;C

(412) & >C◦ * .

(413)

(414) !

(415) D 9 *

(416) . $

(417) 6

(418)

(419) 2;= × ;= 3

(420) “ ” 2(; (3

(421) “ ” 2+; +3 2/

(422)

(423)

(424)

(425) 9

(426) "

(427) 3.

(428) !

(429)

(430)

(431) -

(432)

(433) 2/

(434)

(435)

(436)

(437) 9

(438) " .

(439) 3.

(440)

(441) .

(442) +

(443) .

(444)

(445)

(446) $

(447) +

(448)

(449) ;CC &

(450) @C J ,

(451)

(452) " ) 6 . " *

(453)

(454) DF μ

(455)

(456) 00 )& ! ! &

(457)

(458)

(459) !

(460) D D" 9

(461)

(462)

(463)

(464)

(465)

(466)

(467) "

(468) . ' :DD<"

(469)

(470) /$1

(471)

(472)

(473) +

(474)

(475) . *

(476) "

(477)

(478) "

(479)

(480) 2

(481) D 3

(482)

(483)

(484)

(485) 2

(486) G ?F 3"

(487)

(488)

(489)

(490)

(491)

(492) 0

(493) "

(494)

(495) " 2DC × DC 3 .

(496)

(497)

(498) )

(499)

(500) DC × ;FC ' "

(501) .

(502) ' :DD<"

(503) )

(504)

(505)

(506) .

(507) ,

(508) "

(509)

(510)

(511)

(512)

(513) .

(514)

(515) ,

(516)

(517) "

(518)

(519) "

(520)

(521)

(522) 8 +

(523) &67

(524)

(525)

(526)

(527) +

(528)

(529) .

(530)

(531) 9

(532) 8 + " !

(533)

(534)

(535) -

(536) 2!-3

(537)

(538)

(539)

(540)

(541)

(542) ' "

(543)

(544)

(545)

(546) "

(547) .

(548)

(549)

(550) . D 2 / > 3 ,

(551) 2

(552) / > 3 9

(553)

(554) * " 9. 2

(555) / >3

(556) . . &67

(557)

(558) "

(559)

(560) ) + 9 .

(561) .

(562) 29 3 .

(563) . ,

(564) " 5 9 0

(565) "

(566)

(567)

(568) 2 )

(569)

(570) / >3

(571)

(572) .

(573) B / "

(574) 9

(575) !-

(576)

(577) . ) +

(578)

(579)

(580)

(581) 9 +

(582)

(583)

(584)

(585) + 9 . . * 9 9

(586) * 2 / ? 3 ,

(587)

(588) "

(589) 9

(590) "

(591) 2 / ?3 ! 9

(592) .

(593)

(594)

(595)

(596) 2 / ? 3 4

(597)

(598) / ?"

(599)

(600) 9 "

(601)

(602)

(603)

(604) .

(605) )

(606) "

(607) . "

(608)

(609) )

(610) " ,

(611) 1 $ :=><

(612) * 9 9

(613)

(614)

(615)

(616)

(617)

(618)

(619) 9 + "

(620)

(621) "

(622)

(623) &

(624) "

(625)

(626)

(627) 2 / @3 !-

(628)

(629) .

(630) @C μ

(631) DCC μ

(632)

(633)

(634)

(635)

(636) +

(637)

(638) + .

(639)

(640)

(641)

(642) )

(643)

(644)

(645)

(646)

(647)

(648)

(649)

(650) & *

(651)

(652) 9 *

(653)

(654)

(655)

(656) 2 / G3 +

(657) . . & . .

(658) "

(659)

(660)

(661)

(662) *

(663)

(664)

(665) ! ' :DD<

(666)

(667)

(668) . '

(669)

(670)

(671) "

(672)

(673) 9 ' " *

(674)

(675)

(676) 9 "

(677) 2

(678) 3

(679) + " .

(680)

(681)

(682)

(683) *

(684) ,

(685)

(686) " .

(687) D

(688)

(689) .

(690) 2 / ;C3 ,

(691) . .

(692) "

(693) . +

(694) $ .

(695)

(696) 2 3 +

(697)

(698) &67 :1 = G@ 1 B!! = C F BJ = D B

(699) <

(700) 23

(701)

(702)

(703) 9

(704)

(705) 2/

(706)

(707)

(708)

(709) 9

(710) "

(711) 3.

(712)

(713)

(714)

(715) &67 " 2 3 4 B * 9 " 23

(716) 9

(717) " 23 23

(718)

(719)

(720)

(721) 9 2/

(722)

(723)

(724)

(725) 9

(726) "

(727) 3.

(728) !-

(729) )

(730) &67 2/

(731)

(732)

(733)

(734) 9

(735) "

(736) 3.

(737)

(738) !-

(739) 2 3 9 *

(740)

(741)

(742) 23 *

(743)

(744) &67 2/

(745)

(746)

(747)

(748) 9.

(749) "

(750) 3.

(751) !

(752)

(753) )

(754) . 2 3

(755) 23

(756) :1 = ;F> 1 B!! = C F BJ = = B

(757) < 2/

(758)

(759)

(760)

(761) 9

(762) "

(763) 3.

(764)

(765)

(766)

(767) & .

(768)

(769)

(770) ’ . "

(771) .

(772)

(773) .

(774)

(775)

(776) * .

(777)

(778)

(779) . .

(780)

(781)

(782)

(783)

(784) "

(785)

(786) . "

(787)

(788) &

(789)

(790) &67

(791)

(792)

(793)

(794)

(795)

(796)

(797) & .

(798)

(799)

(800)

(801) "

(802)

(803) .

(804)

(805) 00 78 !$ ! 6 ) 000 + 2

(806)

(807) 3

(808)

(809) *

(810)

(811) 9

(812)

(813) +

(814)

(815)

(816)

(817)

(818) .

(819) I

(820) " . 8

(821)

(822) 2

(823) 3"

(824)

(825) .

(826)

(827) .

(828) * .

(829)

(830) .

(831)

(832) 9 :=>"=?< +

(833)

(834)

(835)

(836) "

(837)

(838)

(839)

(840) 2!-3

(841)

(842) 24!-3 "

(843)

(844)

(845)

(846)

(847)

(848)

(849)

(850) '

(851) " 2

(852)

(853)

(854) !

(855) &3 B

(856)

(857)

(858)

(859) 9 * 2 / ;;3 &"

(860)

(861) . 2

(862) / ;;3

(863)

(864) ;DF μ 2K;DC3

(865)

(866)

(867)

(868)

(869)

(870)

(871) *

(872)

(873) !

(874) . .

(875) / :<

(876)

(877)

(878)

(879) + !-4!-

(880) .

(881) .

(882) 9

(883)

(884) .