METHOD AND APPARATUS FOR MANUFACTURING A MECATRONIC SYSTEM BY THREE-DIMENSIONAL PRINTING

专利摘要:
A method of manufacturing a mechatronic system comprising: - a step of manufacturing a mechanical structure (SM) by three-dimensional printing by deposition of molten wire of at least a first material (M1) electrically insulating; and a step of manufacturing at least one electrical component (CE) in contact with at least one element of said mechanical structure and integral therewith; characterized in that said step of manufacturing at least one electrical component is implemented by three-dimensional printing by deposition of molten wire of at least a second material (M2), conductive or resistive, directly in contact with said element of the structure mechanical. Apparatus for implementing such a method Mechatronic system capable of being manufactured by such a method.
公开号:FR3052698A1
申请号:FR1655552
申请日:2016-06-15
公开日:2017-12-22
发明作者:Mehdi Ammi；Florian Longnos
申请人:Centre National de la Recherche Scientifique CNRS；Universite Paris Sud Paris 11；
IPC主号:

专利说明:

ou P est la résistivité du barreau mesurée dans sa longueur. Ce comportement ohmique est par ailleurs supporté par de nombreuses caractéristiques courant-tension réalisé sur les différents échantillons fabriqués.
La résistance varie également en fonction de la direction d'impression, et d'après la loi ci-dessus, il est possible relier de façon empirique cette variation à la résistivité intrinsèque du motif imprimé, c'est-à-dire à la microstructure de celui-ci. Les résultats obtenus sont reproduits dans le tableau 1 ci-dessous. La résistivité du barreau augmente lorsque la direction d'impression passe de 0° à 45° puis 90°. Cet angle est évalué par rapport à la longueur du barreau.
Tableau 1 - Résistivité des thermoplastiques conducteurs et directions d'impression. L’influence de l’épaisseur « e >> des strates imprimées sur la résistance est en revanche négligeable pour la plupart des matériaux, sauf pour le Matériau 3. La diminution de la résistance d’un facteur 2,5 lorsque e varie entre 100 et 300pm pour le Matériau 3 pourrait s’expliquer par une continuité verticale entre les couches (moins d’interfaces couche-couche homogènes) qui potentiellement permet un plus grand nombre de chemins de percolations. Dans ce cas, l'utilisation d'un recuit local, comme décrit dans la présente invention, permettrait de former de nouveaux chemins de percolation aux interfaces inter-couches et améliorerait significativement la conductivité électrique du motif imprimé.
Afin d’en caractériser la piézorésistivité, les matériaux 1 à 5 ont été utilisés pour fabriquer des éprouvettes de traction uni-axiale en haltère, imprimées par FDM.
Ces éprouvettes ont été conçues en exploitant la différence de résistivité en fonction de la direction d'impression, et avec des dimensions conformes à la norme ASTM D638, c'est-à-dire pour la partie utile : L=50mm, W=10mm, H 300pm. Pour chaque matériau, des exemplaires ont été imprimés entièrement à des directions d’impression, mesurées par rapport à la longueur de l'éprouvette, de 0°, 45° ou 90°. Des exemples ET1 (0°), ET2 (90°) des éprouvettes fabriquées sont illustrés sur les figures 6A et 6B.
La traction a été exercée par le biais de poids suspendus verticalement au bas de l'échantillon, tandis le haut de l'échantillon était fixé via un mors au bâti. Les masses suspendues étaient de 100 g, 200 g, 500 g et 1 kg. En supposant que les éprouvettes n'ont pas de défaut, la contrainte se concentrait dans la partie utile, et la contrainte appliquée variait entre 3 et 33 MPa.
Les résultats des essais sont illustrés sur les figures 7A - 7C.
La figure 7A illustre l’évolution de la résistance électrique R des éprouvettes de traction en Matériau 2 en fonction de la masse M du poids suspendu verticalement. Le taux de remplissage est de 100%, l'éprouvette est donc massive.
La figure 7B illustre l’évolution de la résistance électrique des éprouvettes de traction, sous chargement nul ou de 100g, au cours d'un cyclage mécanique en chargement/déchargement. Les éprouvettes considérées ici ont un taux de remplissage de 80%, ce qui signifie qu’il existe un espacement entre les filaments, ou stries, qui forment le corps de l’éprouvette, dont environ 20% est constitué d’espaces vides.
La figure 7B illustre l’évolution dans le temps t de la résistance électrique des éprouvettes de traction suite au retrait d'un chargement de 100g. Suite à une augmentation brutale, la résistance diminue progressivement vers la valeur de repos.
Parmi les matériaux étudiés, le Matériau 3 a démontré un comportement piézorésistif avec effet de seuil, lorsque l'éprouvette est imprimée à 90°, comme reporté sur la Figure 7A. Le seuil à partir duquel la variation de résistance apparaît peut être abaissé en diminuant le taux de remplissage lors de l'impression par FDM, c'est-à-dire en modulant la qualité du contact entre les stries. Si le taux de remplissage est élevé, il y aura chevauchement entre les stries adjacentes. Plus ce taux diminue, l'épaisseur de l'extrudat étant constant, plus la partie ou il y a chevauchement diminue, et l'interface entre les stries adjacentes voit apparaître des trous ou porosités, où l'air est piégé. Le taux de porosité est donc un paramètre important permettant de contrôler la résistance et le seuil de piézorésistivité du motif imprimé. Après un cycle de déverminage (rupture des chemins faibles qui se produit lors des premières utilisations, conduisant à des variations rapides des propriétés électriques suivies par une phase de stabilisation), on remarque que la variation de résistance apparaît dès 100g de chargement pour un barreau avec un taux de remplissage de 80%, avec une bonne répétabilité et une endurance d'au moins 10 cycles (cf. Figure 7B).
Le même comportement piézorésistif est observé pour les éprouvettes dont les stries sont imprimées à un angle d’impression de 90° par rapport à la longueur de l’éprouvette avec le Matériau 5. Cependant un phénomène de relaxation de la matrice élastomère suite au chargement ou déchargement est observé. Pour le déchargement, ce phénomène induit une augmentation brutale de la résistance puis une diminution logarithmique vers la valeur de repos (cf. Figure 7C). Cette réponse temporelle est ainsi un problème non négligeable dans la fiabilité de la mesure et son exploitation pour réaliser un capteur.
Les modèles comportementaux et lois établis pour la piézorésistivité du Matériau 2, et ceux établis pour la résistivité des thermoplastiques conducteurs permettent de réaliser des capteurs qui tirent partie de ces propriétés.
La figure 8 représente un capteur de flexion bidimensionnel CF2D fonctionnel, entièrement imprimé par FDM avec 3 matériaux thermoplastiques différents. Ce composant comprend en effet une structure SM en thermoplastique isolant (ABS), 4 électrodes de contact ELC1, ELC2, ELC3, ELC4 en thermoplastique conducteur (Matériau 4), et 2 parties centrales JC1, JC2 en thermoplastique chargé ayant un comportement piézorésistif (Matériau 2), imprimé selon des directions d’impression orthogonales, formant des jauges de contrainte, supportées par la structure mécanique SM et dont les axes de sensibilité (déterminés par la direction d’impression, et donc par l’alignement des filaments constitutifs) sons mutuellement perpendiculaires.
Afin d’éviter les temps morts nécessaires pour les changements de bobine, le capteur a été imprimé en utilisant trois têtes d’extrusion, une par matériau. Une fine couche de promoteur d’adhésion a été déposée aux interfaces hétérogènes.
Les parties centrales du capteur sont respectivement un rectangle JC1 de Matériau 2 imprimé à 0° et un rectangle JC2 de Matériau 2 imprimé à 90°, avec une épaisseur de 600pm. D'après les résultats de caractérisation, seul le rectangle dont la direction d'impression est perpendiculaire à la direction de sollicitation en traction produit un changement de résistance électrique. L'autre bloc conserve la même résistance. La juxtaposition de ces deux blocs permet donc de mesurer une contrainte suivant l'axe x ou l'axe y, voire une contrainte bi-axiale en traction.
Lorsque le substrat en ABS subit une contrainte de flexion, sa face supérieure est sollicitée en traction. Cette sollicitation est transmise aux blocs de Matériau 2 par cisaillement à l'interface. Les résultats obtenus avec ce capteur, et démontrant sa fonctionnalité, sont présentés dans le tableau 2 ci-dessous :
Tableau 2.
La liberté de forme permise par le procédé et la disponibilité de thermoplastiques conducteurs et piézorésistifs, permet la réalisation de capteurs de force multidimensionnels. La figure 9F montre une vue latérale d’un tel capteur présentant la forme d’un mini-joystick pouvant transduire l'effort appliqué sur le connecteur central et ses composantes x et y. Ce capteur présente une structure mécanique isolante comprenant un substrat annulaire SA, des piliers-support PS et une plateforme centrale PTC en matériau isolant ; des éléments piézorésistifs PZR1 - PZR4 formant des ponts suspendus réalisés à l’aide d’un support imprimé en polymère sacrificiel, par exemple soluble ; des électrodes de contact EC1 - EC5, le joystick central JC et des piliers conducteurs PC1 - PC4 pour l’accroche de la connectique. Les figures 9A - 9F montrent ces différents éléments séparément.
La figure 10 illustre un capteur acoustique de type piézorésistif, également réalisé par un procédé d’impression selon l’invention. Il comprend une structure mécanique isolante SM, deux électrodes de contact ELC1, ELC2 en thermoplastique conducteur, deux membranes conductrices MCI, MC2, également en thermoplastique conducteur, formant les deux armatures d’un condensateur, et des jauges de contrainte JC1, JC2, reliées chacune à une électrode de contact respective et à la membrane conductrice MCI. Les jauges de contrainte permettent de mesurer les déformations de la membrane MCI sous l’effet d’une onde acoustique.
Comme cela a été évoqué plus haut, une interface de conception est avantageusement prévue pour faciliter la conception des composants électriques imprimés. Cette interface est un système informatique (ordinateur, réseau d’ordinateur, carte à microprocesseur...) programmé pour recevoir en entrée des paramètres d’un composant électrique à fabriquer, tels que les propriétés électriques souhaitées, la position de ses points de prise de contact, son emplacement au sein ou à la surface d’une structure mécanique...et de fournir à sortie, grâce à l’application d’algorithmes appropriés, un fichier d’impression (typiquement au format GCode) qui permet le pilotage de l’imprimante tridimensionnelle.
Les organigrammes des figures 11A à 11H illustrent, à titre d’exemple non limitatif, les algorithmes permettant d’aboutir au fichier d’impression pour un composant électrique simple, à savoir une résistance déposée sur un substrat isolant. La figure 13 illustre, étalement à titre d’exemple non limitatif, un algorithme permettant d’aboutir au fichier d’impression pour une jauge de déformation axiale.
Une résistance fabriquée par impression tridimensionnelle peut être bidimensionnelle (c'est-à-dire planaire, déposée sur une surface) ou tridimensionnelle - passant donc par des points non coplanaires, qui peuvent être imposés. Dans le cas d’une résistance bidimensionnelle, on peut adopter une géométrie linaire - pour des faibles valeurs de résistance - ou en zigzag ou serpentin - pour des valeurs de résistance plus élevés. Ces trois différents cas nécessitent des algorithmes de conception différents. Ainsi, comme illustré dans la figure 11 A, l’utilisateur d’une interface de conception selon l’invention définit d’abord le matériau thermoplastique à utiliser (étape EA1), dont les propriétés sont connues et stockées dans la base de données de modèles BDM, et la valeur de résistance R_user à atteindre (étape EA2), puis (EA3) il indique si la résistance est tridimensionnelle ou doit respecter des points de passage imposés. Dans l’affirmative, l’interface de conception appliquera un algorithme de conception de résistances tridimensionnelles, illustré sur les figures 11G et 11 H. Autrement, le système déterminera de manière autonome (EA4), en fonction de la valeur R_user, s’il convient de réaliser une géométrie linéaire (par exemple si R_user est inférieure à un seuil prédéfini R_th), auquel cas il appliquera l’algorithme illustré sur la figure 11 B, ou bien une géométrie en zigzag (par exemple si R_user est supérieure ou égale à R_th), auquel cas il appliquera l’algorithme illustré sur les figures 11C à 11 F. Le choix entre une géométrie linéaire et en zigzag peut également prendre en compte d’autres paramètres, par exemple d’encombrement, ou être laissé à l’appréciation de l’utilisateur.
Quelle que soit la géométrie choisie, l’algorithme de conception appliqué définit le volume de la résistance à fabriquer (EA5). Ensuite (étape EA6) un modeleur CAO (conception assistée par ordinateur) génère un maillage triangulaire de ce volume sous forme de triangle, qui est typiquement enregistré sous la forme d’un fichier de stéréolithographie au format STL. On procède ensuite (étape EA7) au découpage en tranches (« slicing » en anglais) de volume défini par le fichier STL, ce qui permet de définir chaque tranche de matériau déposée par la tête d’extrusion et d’y associer une trajectoire de déplacement de la tête d’impression. Le découpage est réalisé par un logiciel de type connu en soi, dépendant de l’imprimante, qui de préférence permet également de paramétrer cette dernière en déterminant, notamment, la température des bords de la buse, le taux ou l’angle de remplissage, etc. L’ensemble des déplacements de la buse et des paramètres machines sont enregistrés dans le fichier d’impression Fl, par exemple au format GCode. L’organigramme de la figure 11B illustre l’algorithme de conception d’une résistance linéaire.
Pour commencer, l’utilisateur fournit en entrée au système informatique la largeur Xmax (EB1) et la longueur Ymax (EB2) de la région, supposée rectangulaire, contenant la résistance, la position du point servant d’origine du repère défini dans cette région (EB3), ainsi que les coordonnées des points de contact A et B de la résistance dans ce repère (EB4, EB5) ; ces points sont généralement situés sur des côtés de cette région rectangulaire, en tout cas cela sera toujours le cas dans la suite. Le système informatique calcule alors la longueur L du segment AB (EB6), puis la section S=p L/R_user nécessaire pour obtenir la résistance voulue (EB7), p étant la résistivité du matériau, extraite de la base BDM.
La section S est le produit de la largeur W de la résistance par son épaisseur H, ces deux grandeurs étant à déterminer. L’épaisseur H de la résistance est un multiple entier de l’épaisseur e d’une tranche de matériau déposée par la tête d’extrusion, tandis que la largeur W est comprise entre une valeur minimale Wmin fonction de la tête d’extrusion et la largeur admissible de la résistance, Wmax. On commence en supposant H=e, et on calcule la largeur correspondante : W=S/H. Si on trouve une valeur supérieure à la largeur maximale admissible Wmax, on incrémente H de la valeur e, et ainsi de suite. Bien que cela ne soit pas illustré sur la figure, il est possible que même avec H=e on obtienne W<Wmax. Cela signifie que R_user est trop élevée ; il faut donc soit augmenter la longueur L en déplaçant l’un des points de contact (ou les deux), soit changer de matériau, soit utiliser une géométrie en zigzag. En principe, cette situation ne devrait pas se présenter si la géométrie linéaire a été sélectionnée automatiquement par l’algorithme de la figure 11 A. Ces opérations constituent l’étape EB8.
Ensuite (EB9) on vérifie que la valeur R=RHO L/S soit bien égale à R_user. Si ce n’est pas le cas, l’utilisateur a la possibilité de changer le matériau utilisé, auquel cas l’algorithme reprend à partir de l’étape EB7 ; s’il ne souhaite pas le faire, le système informatique déplace l’un des points A et B (ou les deux) pour modifier la longueur L jusqu’à ce que R=R_user (EB10). L’algorithme de conception d’une résistance en zigzag est illustré sur les figures 11C à 11 F.
Comme dans le cas de la résistance linéaire, l’utilisateur fournit en entrée au système informatique la largeur Wmax (EC1) et la longueur Ymax (EC2) de la région, supposée rectangulaire, contenant la résistance, ainsi que les coordonnées de l’origine du repère (EC3). Il fournit également la largeur W et l’épaisseur H du « fil >> résistif formant le motif en zigzag (EC4, EC5), ce qui permet au système de calculer la longueur de ce fil : L=R_user-W-H/RHO (EC6). Ensuite, l’utilisateur fournit en entrée les coordonnées des points de contact A et B de la résistance (EC7, EC8), ce qui permet au système de déterminer si ces points se trouvent sur des côtés opposés de la région rectangulaire, sur un même côté ou sur des côtés adjacents (et orthogonaux). A ces trois cas de figure correspondent trois algorithmes différents, qui sont illustrés par les figures 11D, 11E et 11F, respectivement.
Lorsque le point B se trouve en face de A (figure 11D), le parcours du fil conducteur est déterminé en partant du point A (étape EDI), en se déplaçant d’un pas prédéfini de longueur DELTA_min (petite par rapport aux dimensions de la région rectangulaire contenant la résistance, mais supérieure à la largeur W du fil) vers le côté où se trouve B et perpendiculairement au côté où se trouve A (ED2). Puis, on se déplace parallèlement au côté de A, dans la direction du sommet de ce côté le plus éloigné de A, en s’arrêtant à une distance DELTA_min du bord de la région rectangulaire, pour garder une marge de sécurité (ED3). Ensuite, on se déplace encore une fois de la longueur DELTA_min vers le côté où se trouve B et perpendiculairement au côté où se trouve A (ED4). La distance parcourue depuis le départ est ensuite mémorisée, et stockée dans une variable dist_p (ED5). A ce point, on calcule la distance minimale qui reste à parcourir, en ligne droite, pour atteindre B (ED6). Si la somme de dist_p et de cette distance minimale dépasse L (calculé à l’étape EC6), cela signifie qu’il n’est pas possible de réaliser une résistance en zigzag de la valeur voulue ; il faut donc changer de matériau. Autrement, le système détermine si les points A et B se font face dans la direction x ou dans la direction y. Dans la suite, on considère seulement le premier cas (étapes ED7, ED8, ED9) ; les opérations effectuées dans le cas opposé sont tout à fait similaires (étapes ED7’, ED8’, ED9’).
La résistance en zigzag est essentiellement constituée d’un certain nombre nb_p de motifs (méandres) formés d’un trait court, d’une longueur d, dans la direction perpendiculaire aux côtés qui portent A et B (direction x, dans l’exemple considéré ici) et d’un trait long, d’une longueur D, dans la direction perpendiculaire, plus deux segments d’extrémité - d’une longueur dist_p du côté de A et d’une longueur à calculer du côté de B. Lors de l’étape ED7, on cherche à résoudre le système : nb_p(d+D)<L-dist_p nb_pd<Xmax-2DELTA_min en faisant une hypothèse sur la valeur de D (car on a trois inconnues et seulement deux inégalités). L’hypothèse initiale est : D=Ymax-2DELTA_min.
Lors de l’étape ED8 on vérifie si la solution trouvée vérifie la condition d>DELTA_min. Si ce n’est pas le cas, la valeur de D est décrémentée de DELTA_min (ED9), et l’étape ED7 est exécutée à nouveau.
Le tronçon final est constitué par le plus court chemin qui relie l’extrémité terminale du dernier motif à B (ED10). En général, pour obtenir une longueur totale égale à L, il sera nécessaire d’ajuster la longueur du dernier trait long (avant-dernier tronçon de la résistance, qui ne sera donc pas nécessairement égale à D. Pour construire le volume de la résistance à partir du parcours ainsi déterminé il ne reste alors qu’à appliquer la largeur du fil W (ED11).
Lorsque le point B se trouve sur le même côté que A (figure 11E), le parcours du fil conducteur est déterminé en partant du point A (étape EE1), en se déplaçant d’un pas prédéfini de longueur DELTA_min (petite par rapport aux dimensions de la région rectangulaire contenant la résistance, mais supérieure à la largeur W du fil) dans la direction perpendiculaire au côté où se trouvent les points A et B (EE2). Puis (EE3) on se déplace parallèlement à ce côté et à l’opposé de B, jusqu’à ne laisser qu’une distance de sécurité DELTA_min du côté perpendiculaire à celui de départ. Puis on se déplace encore de la distance DELTA_min dans la direction du premier déplacement, c'est-à-dire perpendiculairement au côté portant A et B (EE4). Les étapes suivantes - EE7, EE8 à EE10, EE8’ à EE10’, EE11, EE12- sont identiques aux étapes ED6, ED7 à ED9 et ED7’ à ED9’, ED10 et ED11 décrites plus haut.
Lorsque le point B se trouve sur un côté adjacent, et donc perpendiculaire, à celui qui porte le point A (figure 11F), le parcours du fil conducteur est déterminé en partant du point A (étape EF1), en se déplaçant d’un pas prédéfini de longueur DELTA_min (petite par rapport aux dimensions de la région rectangulaire contenant la résistance, mais supérieure à la largeur W du fil) dans la direction perpendiculaire au côté où se trouve le point A (EF2), puis en se déplaçant parallèlement à ce côté et en direction du côté opposé à celui portant le point B, jusqu’à une distance de sécurité DELTA_min de ce dernier (EFS). Ensuite on se déplace parallèlement à ce côté jusqu’à laisser une distance de sécurité DELTA_min du côté opposé à celui de A (EF4). La suite (étapes EFS - EF11) est identique aux cas précédents.
Le dernier cas de figure, celui d’une résistance tridimensionnelle, fait l’objet des figures 11G et 11 Fl. Premièrement, l’utilisateur doit définir un volume dans lequel la résistance est définie, les points de départ (A) et d’arrivée (B), ainsi que des points intermédiaires de passage (EG1). Il est important de noter que, dans la technique d’impression tridimensionnelle FDM, un volume se présente sous la forme d’un empilement de couches d’épaisseur prédéfinie. La position d’un point peut donc être définie en identifiant la couche à laquelle il appartient et en donnant ses coordonnées bidimensionnelles à l’intérieur de cette dernière. Pour définir le parcours de la résistance, on part du point A (EG2), on identifie la projection pt_s du point de passage suivant sur la couche contenant A - ou, plus généralement, le point courant pt_c (EG3), et on détermine la distance à parcourir pour atteindre cette projection (EG4), ainsi que la différence de hauteur entre la couche du point courant et celle du point suivant (EG5). La distance L_cs entre les points pt_c et pt_s est la somme de la distance dans le plan, déterminée à l’étape EG3, et de la différence de hauteur (EG6) ; cette distance est mémorisée. Ensuite, on passe au point suivant, et ainsi de suite, jusqu’à parvenir au point B. On calcule alors L_calc, somme des longueurs L_cs (EG7). La connaissance de la résistivité RFIO du matériau permet alors de calculer la section S du motif résistif (EG8) ; à partir de la connaissance de cette section, l’épaisseur et la largeur sont calculées comme cela a été expliqué en référence au cas linéaire de la figure 11B (étape EG9). On vérifie alors si la résistance ainsi obtenue, R_calc, a la valeur voulue R_user (EG10). Si R_calc, < R_user, et si l’utilisateur ne souhaite pas changer de matériau, il faut remplacer au moins certains des segments droits reliant deux points de passage successifs à l’intérieur d’une même couche par des zigzags. Pour ce faire, on exprime la longueur de chaque segment intermédiaire en pourcentage de L_calc (EG11), et on lui attribue une résistance à atteindre qui correspond au même pourcentage de R_user (EG 12). On procède alors comme dans le cas de la figure 11D - les étapes EG13 - EG16 correspondent aux étapes ED2 - ED5, les étapes EG17, EG18, EG19 aux étapes ED7 à ED9 et ED7’ à ED9’ et les étapes ED20, ED21 aux étapes EDIOetEDII.
Un autre exemple de composant électrique pouvant être fabriqué par impression tridimensionnelle conformément à l’invention, et dont la fabrication peut être facilitée par l’utilisation d’une interface de programmation, est une jauge de déformation (ou contrainte) uni-axiale de type piézorésistif.
La variation relative de résistance ΔΒ d’un élément piézorésistif présentant une résistance au repos Ro vaut : AR/ Ro = k-ε OÙ k est la sensibilité piézorésistive et e la déformation, avec s=AL/Lo, AL étant la variation de longueur et L la longueur au repos de l’élément. Lors de la conception d’un capteur, la longueur (et donc la résistance) initiale est fonction du taux de remplissage du volume du capteur.
Comme tout élément fabriqué par impression tridimensionnelle FDM, un capteur piézorésistif est formé par des filaments, ou bandes; cette structure est illustrée sur la figure 12A. Le capteur fonctionne par le rapprochement ou l’éloignement des bandes sous l’effet d’une contrainte. En effet si l’on rapproche les bandes l’on crée des microcontacts qui font chuter la résistance. A l’inverse, si l’on éloigne les bandes, l’intensité courant qui circule est plus faible. On comprend que la sensibilité aux déformations est beaucoup plus importante pour les déformations parallèles au petit axe de remplissage, indiqué par d sur la figure 12A, que pour les déformations parallèles au grand axe D.
Si le taux de remplissage est égale à 100%, ce qui signifie que les bandes sont parfaitement jointives et en contact mutuel sur toute leur longueur, les variations de résistance ne seront pas mesurables. Il en ira de même dans le cas d’un taux de remplissage trop faible. De manière empirique il a été constaté que des taux de remplissage de 40 à 70% donnent des résultats tout à fait acceptables pour la plupart des matériaux considérés.
Une comparaison des figures 12A et 12B permet de constater qu’un remplissage à 40% (fig. 12A) implique une longueur de filament plus faible qu’avec un remplissage à 70% (fig. 12B). Dans un capteur avec un taux de remplissage de l’ordre de 70%, la conductivité dépend de manière importante de la percolation du courant, c'est-à-dire de la présence de microchemins conducteurs reliant les méandres entre eux. Ces micro-chemins peuvent être facilement interrompus par des faibles déformations de la structure, ce qui fait qu’un capteur ayant un taux de remplissage élevé - mais pas trop, par exemple de l’ordre de 70% - sera capable de détecter de plus petites déformations qu’un capteur avec un taux de remplissage de seulement 40%. Cependant, un tel capteur sera aussi plus sensible aux défauts de fabrication, par exemple, courts-circuits provoqués par une tête d’extrusion sale. A noter que, dans les exemples des figures 12A et 12B, plusieurs couches de matériau résistif ont été déposées, avec des angles de dépôt perpendiculaires, ce qui conduit à une sensibilité piézorésistive isotrope. L’organigramme de la figure 13 illustre les étapes de conception d’un capteur de contrainte piézorésistif. Tout d’abord (étape EPI) L’utilisateur réalise la forme qu’il souhaite sur un logiciel de CAO et l’importe en format STL (maillage avec des triangles du volume). Ensuite, afin d’être sûr que le capteur sera suffisamment souple et résistant, on fixe sa hauteur H à une valeur prédéfinie, par exemple trois fois l’épaisseur d’impression e (EP2).
La dernière couche servira de capteur. Les autres couches peuvent par exemple être réalisées en matériau isolant avec un remplissage fixe, généralement compris entre 60 et 100% et par exemple égal à 70%, de manière à laisser plus ou moins de souplesse au capteur (EPS). En variante, des couches conductrices ou résistives présentant une autre fonction peuvent également être présentes au-dessous de celle servant de capteur de contrainte.
Sur la couche du capteur, l’utilisateur définit la position des électrodes (EP4) et choisi le matériau résistif, parmi une liste qui lui est proposée (EPS), ainsi que la sensibilité voulue - également parmi une liste de valeurs possibles (EP6). Le remplissage se fait généralement en mode rectiligne (EP7), avec une direction principale de remplissage perpendiculaire à l’axe défini par les deux électrodes (EPS).
Un tableau, construit lors d’une étape préalable d’étalonnage et enregistré dans une mémoire de l’ordinateur exécutant le procédé, permet de trouver le taux de remplissage permettant d’atteindre la sensibilité voulue, étant donnés le matériau choisi et la géométrie voulue (EP9).
La vitesse d’impression (EP10) peut être déterminée en fonction du taux de remplissage également déterminé, typiquement au moyen d’un autre tableau de correspondance construit par étalonnage. Par exemple, on peut choisir une vitesse d’impression faible pour des taux de remplissage importants afin d’éviter les courts-circuits qui peuvent se produire lorsque des résidus de filament restent collés à la buse et se décollent de manière impromptue durant l’impression. Pour des taux de remplissage plus faibles, l’écartement des lignes de passage, et donc l’absence de matière dans le passage de la buse, suffit généralement à éviter qu’un résidu vienne se coller ; on peut donc imprimer à une vitesse plus élevée, par exemple de 60 mm/s. En variante, on peut utiliser une vitesse d’impression relativement faible (par exemple 30 mm/s) indépendamment du taux de remplissage. L’invention a été décrite en référence à des systèmes mécatroniques comportant des capteurs piézorésistifs. Cependant, l’invention n’est pas limitée à ce cas de figure. Il est en effet possible, en particulier en choisissant de manière opportune les charges contenues dans la matrice thermoplastique, d’imprimer par FLM des matériaux sensibles à la température, à la lumière, aux agents chimiques, etc. de manière à obtenir des capteurs thermiques, optiques, chimiques, respectivement. Par ailleurs, à partir de matériaux conducteurs et isolant, il est possible de fabriquer des capteurs capacitifs ou inductifs, et des antennes utilisables en émission et/ou en réception. Il est également possible de réaliser des actionneurs thermiques, électrostatiques, magnétiques (en utilisant des charges ferromagnétiques) ou même piézoélectriques. Il est également possible d’imprimer des OLED (diodes électroluminescentes organiques) et/ou des cellules photovoltaïques organiques.
Dans certains cas, l’invention permettra la réalisation d’un système mécatronique entièrement imprimé. Dans d’autres cas, seule une partie du sous-ensemble électrique/électronique d’un tel système sera imprimée (capteurs, lignes conductrices, actionneurs...) tandis que d’autres composants - par exemple des circuits intégrés réalisant des fonctions électroniques complexes, pourront être rapportés. Même dans ce dernier cas, l’invention permet de réduire le nombre d’opérations d’assemblage, et donc de rendre la fabrication du système mécatronique plus rapide et moins onéreuse. Références [Coiai, 2015] Serena Coiai et. Al, « Nanocomposites Based on Thermoplastic Polymers and Functional Nanofiller for Sensor Applications >>, Materials, vol. 8, pages 3377-3427, 2015.
[Deng, 2014] Hua Deng et al., « Progress on the morphological control of conductive network in conductive polymer composites and the use as electroactive multifunctional materials », Progress in Polymer Science, vol 39, pages 627-655, 2014.
[Muth, 2014] Joseph T. Muth et al., « Embedded 3D Printing of Strain Sensors within Highiy Stretchable Elastomers », Advanced Materials, vol 26, pages 6307-6312, 2014.
[Sitthi-Amorn, 2015] Pitchaya Sitthi-Amorn et al. « MultiFab: A Machine Vision Assisted Platform for Multi-material 3D Printing » ACM Transactions on Graphics (SIGGRAPH 2015).
[Rossiter, 2009] Jonathan Rossiter et al. « Printing 3D dielectric elastomer actuators for soft robotics » Proc, of SPIE Vol. 7287, 72870H (2009).
[Leigh, 2012] Simon J. Leigh et al. “A Simple, Low-Cost Conductive Composite Matériel for 3D Printing of Electronic Sensors”, PLOS ONE, Vol. 7, Issuer 11, e49365 (Nov. 2012).
[O’Brien, 2015] Jonathan O’Brien et al. « Miniaturization of Microwave Components and Antennes Using 3D Manufacturing »9th Européen Conférence on Antennes and Propagation (EuCAP), Lisbonne, ΙΟΙ 7 mai 2015.
The invention relates to a process for the manufacture of a mechatronic system, to a mechatronic system obtainable by means of such a method, and to a device for the manufacture of a mechatronic system by three-dimensional printing. adapted to implement such a method. The invention is based on three-dimensional printing techniques (3D), also known as additive manufacturing. It lends itself to many applications, such as the manufacture of: physiological sensors or portable activity (heart rate, blood glucose ...); rehabilitation devices; two- or three-dimensional interaction interfaces; connected objects; etc.
Today, additive manufacturing techniques (or 3D printing) are booming in different areas and are poised to revolutionize the industrial sector, but also the consumer mode of consumption. Various processes have been developed in recent years to design a wide variety of objects and mechanical structures. They make it possible to control the local or global mechanical properties or the appearance of the objects (eg color or texture). These processes, however, only produce passive objects, with no capacity to perceive or act on the environment.
To make these objects active, it is necessary to integrate electronic components and functions, which at present are manufactured separately, by essentially subtractive methods, then assembled with a mechanical structure in industrial chains for the production of an end object. . Integrating electronic functions into a non-planar, non-assembly mechanical structure is a challenge.
In recent years, many actors have been interested in the use of polymer materials for the production of so-called flexible or organic electronics. This branch of electronics is relatively recent since the first conductive polymers were developed in 1977 and the first electronic components using these materials were born in the mid-1980s. Today organic electronics can realize numerous electronic components such as field effect transistors (OFET), photovoltaic cells (OPV), organic light-emitting diodes (OLEDs), electrochemical biosensors or even electroactive polymer-based actuators (EAPS). To realize this electronics new processes have been developed, and some are now used on an industrial scale such as continuous or rotary web printing (flexography, rotogravure, ...). Ink jet (IJP) and aerosol jet (AJP) printing, which are part of the 3D printing processes, make it possible to produce certain organic electronic components, including sensors [Muth, 2014; Sitthi-Amorn, 2015] and are currently the subject of intensive research on this subject.
However, these processes produce only components on planar substrates [Rossiter, 2009] or which require additional assembly operations, thus not allowing the design of complete mechatronic 3D structures. Recently, it has been proposed to use conductive polymers with the Fused Deposition Modeling (FDM) 3D printing method for printing pressure sensors on 3D objects [Leigh , 2012]. It has also been proposed to use functional thermoplastics for printing elementary components such as antennas [O'Brien, 2015].
The potential of polymer matrix composites for many applications such as sensors and actuators has already been demonstrated: [Coiai, 2015], [Deng, 2014]. In addition, the science of nanocomposites and nanoparticles (cationic nano-clays, anionic nanoclays, noble metal nanoparticles, carbon nanotubes, etc.) has led to a better understanding and control of the synthesis processes of these materials. In particular thermoplastics incorporating carbonaceous fillers are promising for the manufacture of sensors, for example deformation, force, temperature, electrochemical (liquid or gas detection), etc. Composites with a metal core (for example in copper) and a thermoplastic matrix, developed and integrated on an FDM printer, make it possible to produce strain gauges, are also known (see US 2014/328964). At present, however, it is not possible to manufacture complete mechatronic systems by 3D printing. Hybrid processes now allow this, but with limited potential. The most successful of these is the VoxelS platform (www.voxel8.co), a spin-off of Harvard University, which combines the FDM for the structure of the object, a metallic ink for conductive tracks of the circuit, and a discrete component positioning system. The invention aims to overcome the aforementioned drawbacks and limitations of the prior art. More particularly, it aims to allow the manufacture, in one piece and by the same 3D printing process, the mechanical subsystem and at least a portion of the electrical subsystem of a mechatronic system. The term "electrical" should be interpreted broadly, including electronic and / or electromagnetic or even optoelectronic functionalities.
According to the invention, this object is achieved by using the three-dimensional printing technique by deposition of molten wire to manufacture both a mechanical structure and at least one electrical component (conductive track, resistance, sensor ...) integral with said structure, for example arranged on its surface. This requires the use of at least two different materials: a first material, electrically insulating, used to print the structure, and at least a second material, conductive or resistive, used to print the electrical component or components. This avoids the disadvantages of hybrid techniques: additional cost due to assembly steps, mechanical fragility, size ...
An object of the invention is therefore a method of manufacturing a mechatronic system comprising: a step of manufacturing a mechanical structure by three-dimensional printing by deposition of molten wire of at least a first electrically insulating material; and a step of manufacturing at least one electrical component in contact with at least one element of said mechanical structure and integral therewith; characterized in that said step of manufacturing at least one electrical component is carried out by three-dimensional printing by deposition of molten wire of at least a second material, conductive or resistive, directly in contact with said element of the mechanical structure.
According to particular embodiments of such a method: Said or said electrical component may be a transducer, and more particularly a piezoresistive sensor.
Said second material, conductive or resistive, may comprise conductive fillers dispersed in a thermoplastic insulating matrix.
The method may also comprise a local annealing step implemented during or after the deposition of a layer of the first or second material, in correspondence of said deposit.
The method may also comprise a step of depositing an adhesion promoter on a surface of the mechatronic system during manufacture before depositing, above said surface, a layer of a different material.
The method may include the use of at least two separate extrusion heads for depositing the first and second material.
The method may also include a step of generating a print file for producing at least one said electrical component, said step being computer-implemented and comprising: a substep of providing said computer with data indicative of a position of one or more points of contact, a spatial region where said component is to be manufactured and at least one electrical property of said component; a sub-step of calculating a geometry of said component by applying to said data a predefined mathematical model; and a sub-step of generating said print file making it possible to produce said geometry by three-dimensional printing by deposition of said or said at least one second conductive or resistive material.
Another object of the invention is a mechatronic system comprising an electrically insulating mechanical structure and at least one electrical component arranged in contact with at least one element of said mechanical structure and integral with it, characterized in that the mechanical structure and the component electric are made in one piece by three-dimensional printing of at least a first electrically insulating material forming said mechanical structure and at least a second conductive or resistive material forming said electrical component.
Yet another object of the invention is an apparatus for carrying out a method as mentioned above, comprising a three-dimensional printer of the molten wire deposition type having at least two separate, activatable extrusion heads. independently and adapted to deposit two different materials, said extrusion heads being arranged side by side with the same extrusion direction and being carried by a same printing carriage ensuring their simultaneous movement, the print head also comprising a mechanism for moving an inactive extrusion head in a direction opposite to said extrusion direction when said one or said other extrusion head is active.
According to particular embodiments of such an apparatus: said print carriage may be equipped with a capacitive sensor configured to measure its distance from a printing surface. In this case, the apparatus may also comprise a printing plate above which said printing carriage moves, said plate being equipped with metal electrodes that can be detected by said capacitive sensor, whereby said sensor enables a calibration of the position of the print carriage with respect to the tray. The apparatus may be equipped with: a camera configured to acquire an image of a layer of material deposited by the three-dimensional printer; an image processing system configured to compare said image with a three-dimensional model stored in a computer memory and derive a position error from said print carriage; and a computer system for controlling said carriage configured to correct said position error when depositing a successive layer of material. The apparatus may also include a computer system for generating a print file for driving said three-dimensional printer so as to produce an electrical component, said computer system being configured to: input data indicative of a position of one or more contact points, a spatial region where said component is to be manufactured and at least one electrical property of said component; computing a geometry of said component by applying to said data a predefined mathematical model; and generating a print file for making said geometry by three-dimensional printing by depositing molten wire of at least one conductive or resistive material. The apparatus may also include an electromagnetic radiation beam generator, said generator being configured to produce local heating of a material deposited on a printing surface of said three-dimensional printer. Other characteristics, details and advantages of the invention will emerge on reading the description made with reference to the appended drawings given by way of example and which represent, respectively: FIG. 1, a block diagram of an apparatus according to a embodiment of the invention; FIGS. 2A-2C, a schematic representation of a local annealing step of a manufacturing method according to one embodiment of the invention; Figure 3, a spray nozzle of an adhesion promoter used in a manufacturing method according to one embodiment of the invention; 4A - 4E, various conductive patterns made by a manufacturing method according to one embodiment of the invention; FIGS. 5A-5C, graphs illustrating the evolution of the resistance of different resistive patterns realized by a manufacturing method according to an embodiment of the invention as a function of their geometrical parameters; FIGS. 6A and 6B, traction specimens equipped with uniaxial strain gauges, made by a manufacturing method according to one embodiment of the invention; FIGS. 7A-7C, graphs illustrating the results of measurements made during tensile tests of specimens of the type illustrated in FIGS. 6A and 6B; FIG. 8, a two-dimensional bending sensor produced by a manufacturing method according to one embodiment of the invention; FIGS. 9A-9E, the manufacturing sequence of a multi-axial force sensor produced by a manufacturing method according to one embodiment of the invention, and FIG. 9F a side view of this sensor; Figure 10, a sectional view of an acoustic sensor produced by a manufacturing method according to one embodiment of the invention; FIGS. 11A to 11H, flowcharts illustrating an algorithm for generating a print file for an electrical resistance produced by a manufacturing method according to one embodiment of the invention; Figures 12A-12C, illustrations of the concept of filling a surface; and FIG. 13, a flowchart illustrating an algorithm for generating a print file for a piezoresistive strain gauge produced by a manufacturing method according to an embodiment of the invention.
Figure 1 schematically illustrates an apparatus for manufacturing three-dimensional mechatronic objects in accordance with one embodiment of the invention. These objects, manufactured by deposition of molten filament and in particular of functional thermoplastics, comprise a mechanical structure, possibly articulated, and electronic components such as passive components (resistor, capacitor, antenna, etc.), sensors and actuators. The apparatus of Figure 1 makes it possible to print complex 3D objects with capabilities of interactions with their environment or the user, with a short design and manufacturing cycle, reduced costs, great flexibility in the form of objects and their use. The apparatus essentially comprises a three-dimensional printer I3D, of the type of filament deposition - or FDM - adapted for the implementation of a method according to the invention. It may also include a printed electrical / electronic components design interface. This interface may include a BDM database, stored on a computer readable medium such as a hard disk, containing a library of electronic components and transducers made by three-dimensional printing using functional thermoplastics, with their behavioral models. It may also comprise an SGF computer system (typically a computer) configured to receive, via a user terminal and possibly a graphical interface, the desired electrical and geometric properties of an electrical component and exploiting them. database models for generating a print file Fl containing all the instructions that allow the I3D printer to manufacture the electrical device so designed. The steps leading to the generation of the file Fl will be described later in this document, with reference to FIGS. 11A-11H and 13.
As represented in FIG. 1, the three-dimensional printer I3D comprises at least two extrusion heads TE1, TE2 arranged side by side, carried by the same printing carriage C1, movable in three orthogonal directions, x, y and z - the direction z corresponding to the extrusion direction of the two extrusion heads - thanks to a displacement mechanism MD, whose structure is not shown in detail because it is conventional, driven by a computer system (computer or card to microcontroller) SIP. The use of two separate extrusion heads facilitates the deposition of at least two different thermoplastic materials: an insulating material M1 used for the manufacture of the mechanical structure SM of the mechatronic system, and a functional material (conductive or resistive) M2 used for the manufacture of at least one electrical component CE. The nature of these materials will be discussed in detail below. In general, it will most often be thermoplastic polymer materials or composites having a thermoplastic polymer matrix. At least the material M2 will generally contain charges influencing its electrical properties, for example by making it conductive.
For the realization of complex systems, comprising more than two materials, it will be possible to use more than two heads; the same head can be used to deposit several different materials, but this slows down the process (it is necessary to make changes of feed material of the nozzle) and introduces a risk of contamination.
An ATV actuating mechanism makes it possible to adjust the relative vertical position (in z) of each head relative to the other heads. In particular, this makes it possible to elevate the inactive heads during the production of complex structures, where the risk of collision between the extrusion heads and the elements already printed becomes important. Even in the case of a single layer-by-layer printing, residues on the inactive heads are deposited involuntarily and uncontrollably, which can alter the aesthetic (color, texture, etc.) and, above all, functional properties. of the final object. For example, this can lead to short circuits between conductive tracks. In a conventional manner, each extrusion head TE1, TE2 comprises an extrusion nozzle having heating edges and a coil leading to this nozzle a filament of the thermoplastic material M1, M2 to be deposited. Typically, each nozzle is mounted on a heating block, which transmits heat to the extrusion edges by thermal conduction. The end of the filament in contact with the heating edges of the nozzle melts, and the molten material is ejected from the nozzle under the effect of the pressure exerted by the not yet melted portion of the filament, which acts as a piston.
The control of the amount of material by the printer, can be achieved by a simple weight control, knowing the density of the filament, the weight of the support coil (standard, but can be deduced from its dimensions and the density of the material constituent) and the diameter of the filament. The weight can be measured by a simple pressure sensor, or a more complex force sensor, which could come from 3D printing itself.
For the sake of accuracy, real-time measurement of weight by the sensor can be supplemented by a more conventional approach of using a contact sensor to measure the number of rotations made by the coil. Each turn a counter is incremented by 1. The length consumed for a complete turn is equal to the perimeter of the coil; by multiplying this length by the section of the filament and by its density, the quantity of deposited material is obtained.
The reliability of the printing of multi-material and functional structures depends in particular on the accuracy of the positioning of the different extrusion heads. The reliability relies firstly on an automated calibration procedure for the spacing between the extrusion heads and the PLI printing plate (in the z-direction), and for the position of the heads in the plane (x, y) the plateau. Calibration of the spacing between the extrusion heads and the height plate (z), and the flatness of the plate, is generally done by means of a limit switch and a setting of the corners of the plate. But when using large surface trays (for example of the order of 20 cm X 20 cm or more), it becomes difficult to ensure its flatness and therefore the use of a limit sensor n ' is not satisfactory. This is why an apparatus according to one embodiment of the invention preferably comprises a capacitive sensor DC integral with the printing carriage. The use of the capacitive sensor makes it possible to calibrate the plate at several points without touching it and to avoid the mechanical inaccuracies inherent in the mechanical end-limit sensors commonly used on the servomotors. The capacitive sensor acts as a non-contact switch. Unlike an inductive sensor, it detects non-ferrous materials such as glass, wood, skin, etc. This sensor will actually simply replace the limit switch being installed directly on the print carriage. The calibration of the position of the extrusion heads in the xy plane of the plate is important in order to align the structural parts made with the different heads.
For a first layer (realizing, for example, a surface of the mechanical structure), an initial calibration is carried out thanks to the capacitive sensor CC with integrated ground electrode, and metal electrodes of mass (references EET1, EET2 in FIG. ) placed in the corners of the board. Indeed, at a constant distance between the sensor and the plate, the measured capacitance will be greater with a metal piece between the two electrodes of the sensor rather than with the coating or insulating material of the plate.
For subsequent layers, especially when multiple print heads are used, and thus more than one thermoplastic material is printed, alignment with the previous layer and with the printed part with another material should be checked with the aid of a visual servoing using a CE camera connected to a STI image processing system (which, in the embodiment of Figure 1, coincides with the SIP computer control system). The visual servoing is carried out as follows: an image of the printing surface, on which at least one layer of material has been deposited, is acquired by the camera CE; optionally, multiple images can be acquired and averaged to improve the signal-to-noise difference; a reference point is chosen on the image, automatically or manually (intervention of a user); the position of this reference point on the image taken by the camera is compared to its position on the model of the last layer deposited, already in memory in the printer; the position of the print carriage along the x and y axes is calculated, and stored in memory; the printing of the upper layer takes into account these possible offsets.
Preferably, as in the case of Figure 1, the calibration camera CE is secured to the carriage, with a relative position relative to the extrusion heads which is known. Alternatively, the camera may be stationary and use a visual cue attached to the print carriage to determine the relative position of the latter relative to the reference point on the image.
A process according to the invention necessarily involves at least two different materials deposited in superposed layers. However, the adhesion between heterogeneous layers is not always satisfactory. For this reason, according to a preferred embodiment of the invention local in situ annealing is performed at the same time as, or after, the deposition of a layer. The role of this local annealing is to provide thermal energy to enhance the adhesion between two materials constituting the object; incidentally it may also improve the intrinsic properties of the functional thermoplastic materials used.
Current FDM technology is based on the simple extrusion of adjacent polymer layers or on top of each other, and uses latent heat of extrusion, with or without the aid of a hot plate, to melt and weld the layers. contiguous. This process induces an incomplete or non-uniform welding of the layers, and consequently reduces the mechanical properties due to a delamination between the layers, in particular in the case of a load applied perpendicular to the surface of the layer or the direction of Filament deposition. This is all the more true for the interfaces between two different materials, where the wetting between the two materials can be partial. During the deposition of the filament of a second material on a layer of a first material, the filament of the second material is in the molten state and the wetting of the surface of the layer of the first material depends on physicochemical parameters of the two materials , such as the surface tensions of the two species, the surface roughness of the first material and the viscosity of the second material. During manufacture and then cooling of the part, the delamination of the layers can then take place due to reduced contact surfaces, adhesion forces that are too low, and differences between the thermal expansion coefficients of the two materials.
The annealing carried out during the implementation of the invention is based on the local heating of the materials deposited by means of nanofillers (CM1, CM2 in FIGS. 2B and 2C), incorporated in the matrices of the materials (MM1, MM2 on these same figures) and capable of absorbing an electromagnetic beam irradiating at one or more given wavelengths. In the embodiment of FIG. 1, an electromagnetic radiation source SRR, for example an infrared laser, is attached to the printing carriage; the reference FL designates the electromagnetic beam generated by the source (here a laser beam, but it could also be another type of radiation, for example microwaves, see for example WO2015130401, or radio frequencies).
The annealing can be carried out during the printing of the layer in question, according to different configurations: passage of the beam behind the active extrusion head and action on the last layer deposited (configuration shown in FIG 2A, where the reference ΔΤ represents symbolically localized heating); passage of the beam independently (inactive extrusion head), with an action on: the last layer or layers deposited in the case of a homogeneous material; the last layer or layers of the first material, located below the second, by transmission therethrough, which is transparent to the wavelength used.
The annealing system has been very schematically represented in FIGS. 1 and 2A; in practice it will include a radiation emitting source, a focusing system (possibly integrated with the source) and a system for guiding the beam to the point of impact, typically an optical fiber. In the embodiment of Figure 2A, the annealing radiation source SRR is integral with the print carriage, but in other embodiments it may be independent.
Radiation-absorbing nano-fillers can be selected from the group consisting of carbon nanotubes, carbon black, buckyballs, graphene, supermagnetic nanoparticles, magnetic nanoparticles, metallic nanowires, semiconductor nanowires, quantum dots, polyaniline (PANI), poly3,4 -ethylenedioxythiophene polystyrenesulfonate, and combinations thereof. Their choice will be made not only on the basis of their functionality thus transferred to the composite (for example an electrical conductivity) but also on the basis of the wavelength of the radiation corresponding to the maximum optical absorbance of the load and therefore optimal heating for minimal power. For example, many fillers have a high absorbance to microwaves, for example metals, oxides, carbon (especially carbon nanotubes "CNT") and conductive polymers (eg polypyrrole). The use of radiofrequency radiation (of the order of ten MHz) is adapted for ceramic charges, such as SiC, ZnO or TiO 2. The use of an infrared laser is particularly effective for uncharged or fiber loaded matrices (eg carbon fibers). The radiation absorbing charges may already be present in the molten filament, or may be deposited by spraying a suspension.
The annealing allows the interdiffusion of the polymeric chains of the matrices between the adjacent layers (FIG. 2C), while limiting the heating of the rest of the part and thus significantly reducing the alteration of its dimensions, in particular by creep and thermal cycling.
The distribution and distribution of CM2 functionalising fillers in the deposited filaments can be modified during the extrusion process, or not initially optimal. Local in-situ annealing also makes it possible to homogenize the distribution of the charges in the volume of the matrix, to reform a maximum of percolation networks (conductor, semiconductor or dielectric according to the nature of the charges), and therefore of improve the functional properties of the composite in the final structure. This is schematically illustrated in Figure 2B.
The deposition of an adhesion promoter at the interface between two parts made of different thermoplastics, and in particular with different matrices in the case of composites, can be used as a replacement or complement for irradiating beam annealing and to improve the between those two parties. This deposition is performed by means of a droplet spray nozzle, for example a valve with atomizing air, air for control and liquid adhesion promoter. Such a device is shown very schematically in FIG. 3, where it is identified by the reference BP; APA refers to the adhesion promoter agent stream.
It is known that the spray deposition pattern of the adhesion promoter on the surface is Gaussian and centered on the point opposite the sprayer. In order to obtain a concentration profile as uniform as possible it is therefore necessary to program passageways for the spray nozzle. The adhesion promoter may be a homogeneous liquid or a suspension of nano or micron particles. In the latter case, it will be necessary to make sure to have a homogeneous suspension beforehand during its preparation, in particular by means of sonication and centrifugation steps. In all cases, the reaction of the adhesion promoter with the interfaces and / or the vaporization of the solvent must allow to leave in fine a solid adhesion layer. In addition, it is expected that the spraying device can allow to deposit absorbent particles, which once sandwiched between two layers, a same thermoplastic or two different thermoplastics can be irradiated with the beam described above and allow improve membership in this way.
As mentioned above, the structural material M1 of a mechatronic object according to the invention is an electrical insulating material. The structural material may be a polymer, such as a thermoplastic, a thermoplastic elastomer, a ceramic, or a thermoplastic or ceramic matrix composite. The thermoplastic matrix may be part of the group, but not only, including acrylonitrile butadiene styrene (ABS), polylactic acid (PLA), polyamide (nylon), polyimide (PI), polyethylene (PE), polypropylene (PP), polystyrene (PS), polytetrafluoroethylene (PTFE), polyvinylchloride (PVC), polyurethane (PU), polycarbonate (PC), polyphenylsylfone (PPSU), polyether ether ketone (PEEK), and mixtures thereof. The ceramic may be selected from the group consisting of oxides, carbides, borides, nitrides and silicides. For example, compatible ceramics include silicon nitride, PZT, aluminum oxide or hydroxypatite. The composites used can incorporate any type of charges, which make it possible to modify and adjust its mechanical and thermal properties, such as ceramic or metallic fillers, glass or carbon fibers, or carbonaceous particles.
The electrical component or components are made of at least one functional thermoplastic material M2, which is generally a composite having properties such as good electrical conductivity and / or piezoresistivity or piezoelectricity and / or good thermal conductivity and / or a coefficient high dielectric, etc. For example, for an electrically conductive thermoplastic, the charges may be part of the group of carbonaceous particles including carbon black, graphene, carbon nanotubes. To obtain a good thermal conductivity. metal fillers or nanotubes may be incorporated. This material should have the proper properties for FDM shaping, namely a melting temperature close to that of the structural material (s) and less than 300%, and a sufficiently low viscosity in the melting zone. For this, a solution is, as far as possible, to use a polymer matrix identical or similar to that of the material of structure M1. Used alone or in combination with another thermoplastic, functional or otherwise, the functional thermoplastic M2 is used in the manufacture of an electrical or electronic component such as a passive component (resistor, capacitor, antenna, etc.), a sensor or an actuator .
To take full advantage of the advantageous characteristics of an apparatus and a method according to the invention, it is possible to develop and integrate new thermoplastic materials having functional properties for producing electronic components, sensors and actuators. The integration of these materials requires exhaustive electrical and mechanical characterizations, in order to identify the parameters to be adjusted to guarantee optimal functionality in the 3D Mechatronic object. For this, the inventors are particularly interested in the following composite thermoplastics:
material 1: ABS matrix / CB fillers (carbon black) material 2: PLA matrix / CB fillers material 3: PI-ETPU matrix (polyimide / engineering thermolastic polyurethane) / CB fillers material 4: PLA matrix / CNT fillers (carbon nanotubes) ) material 5: PLA matrix / graphene fillers The materials that were the subject of these characterization studies are thermoplastics loaded with carbonaceous particles having a high intrinsic electrical conductivity. These composites have an electrical conductivity which depends not only on the nature of the charges, but also their concentration and distribution in the thermoplastic matrix. The concentration of the charges must be sufficient to allow the formation of conduction or percolation path, but low enough that the viscosity of the filament allows the continuous extrusion of the molten filament in a temperature range below 300 ° C. The distribution of the charges in the composite filament depends on the process of shaping the composite filament (single or twin screw extrusion), the control of the process parameters and the preparation of the particles with a potential surface functionalization. A homogeneous distribution and distribution of the charges allows a greater number of percolation paths and thus a better electrical conductivity of the composite. But in fine the electrical properties of the composite are also modulated also by the process and geometric parameters used during the printing of the 3D object by FDM.
As indicated above, the viscosity, flexibility and hardness of the five composites studied is not the same because of their matrices, their charges and their shaping which differ. For this, the process parameters must be adapted to allow continuous extrusion of the filament. For the materials studied, the extrusion temperature depends mainly on the thermoplastic matrix used: 210-220 ° C for the charged PLA, 230 ° C for the loaded ABS and 220 ° C for the loaded PI-ETPU. It is clear however that these values have been optimized by the manufacturers, and in particular by adopting the appropriate concentration of charges.
The speed of movement of the nozzles during printing, or printing speed, is governed by parameters other than viscosity, however. The extrusion of the filament is allowed thanks to the meshing of the filament, pinched between a fixed wheel and a moving wheel, and the upper part of the filament, solid, is piston on the lower part, liquid. The flexible filaments have a tendency to buckle under the effect of compressive stress, and this despite an extrusion head design to limit the lateral movements of the filament. Lowering the velocity, typically 20mm / s, allows time for the molten filament to be extruded, and to remain below the limiting creep stress for the flexible filaments.
Carbon nanoparticles, for example CNTs and graphene, are not only good electrical conductors, but also good thermal conductors. A print speed that is too low may allow the core of the molten filament to cool down by internal heat conduction, despite contact at the edges with the hot nozzles. This phenomenon requires limiting the transit time of the filament in the nozzle, so that it is long enough to allow liquefaction of the filament, and short enough to prevent the resolidification of the core. In this case, a high printing speed, typically 80mm / s, allows continuous extrusion and avoid clogging (in English) of the nozzle.
A final parameter to take into account depending on the nature of the thermoplastic matrix, and its behavior in the molten state, is the distance between the nozzle and the tray for the first layer. Indeed, a high viscosity and elasticity require a greater spacing, as for ABS compared to PLA in the first case, and for PI-ETPU compared to PLA in the second case.
To make mathematical models of electrical components that can be manufactured in accordance with the invention, rectangular bar-shaped patterns have been made from the above five materials, and characterized. These patterns differ in their length L (FIG 4A), width W (FIG 4B), overall thickness H (FIG 4C), thickness e of each elementary layer, or "stratum" (FIG 4D) and direction of printing (Figure 4E). Extensive electrical characterization has made it possible to classify the materials according to their conductivity and to establish laws linking the resistance of the patterns to the geometric parameters mentioned above.
FIG. 5A shows the evolution of the resistance of a rectangular bar (W = 20 mm, H = 400 μm, print direction with respect to the length of the bar = 0 °, T = 200 μm) as a function of its length L for the 5 materials studied. The most conductive materials are easily identifiable: the conductivity decreases when the index of the material increases from 1 to 5. Figure 5B shows the evolution of the average relative resistance of the bars (L = 50mm, H = 0.4mm, direction of printing with respect to the length of the bar = 0 °, T = 200 μm) as a function of its width for the 5 materials studied. The average relative resistance is given by (R-R0) / R0, where R is the resistance of the bar considered and RO that of a reference bar having a length equal to 10 mm. The evolution in inverse law is valid for all, demonstrating their ohmic behavior. FIG. 5C shows the evolution of the average relative strength of the bars as a function of their width (L = 50 mm, H = 0.4 mm, printing direction relative to the length of the bar = 0 °, T = 200 μm) and of their height (L = 50mm, W = 20mm, printing direction relative to the length of the bar = 0 °, T = 200pm) for Material 4. The resistance is inversely proportional to the width and height of the bar , with the unit power.
The relative resistance variation R of the patterns as a function of length L, width W and height H is almost identical for the five composites. Qualitatively, in Figures 5A-G, it is also noted that once printed thermoplastics behave as ohmic conductors, and that the resistance obeys the general expression:
or P is the resistivity of the bar measured in its length. This ohmic behavior is also supported by many current-voltage characteristics made on the various samples manufactured.
The resistance also varies according to the direction of printing, and according to the above law, it is possible to empirically connect this variation to the intrinsic resistivity of the printed pattern, that is to say, the microstructure of it. The results obtained are reproduced in Table 1 below. The resistivity of the bar increases when the printing direction goes from 0 ° to 45 ° then 90 °. This angle is evaluated relative to the length of the bar.
Table 1 - Resistivity of conductive thermoplastics and printing directions. The influence of the thickness "e" of the strata printed on the resistance is on the other hand negligible for most materials, except for the Material 3. The decrease of the resistance by a factor 2.5 when e varies between 100 and 300pm for Material 3 could be explained by a vertical continuity between layers (fewer homogeneous layer-layer interfaces) that potentially allows for a greater number of percolation paths. In this case, the use of local annealing, as described in the present invention, would make it possible to form new percolation paths at the interlayer interfaces and would significantly improve the electrical conductivity of the printed pattern.
In order to characterize the piezoresistivity, materials 1 to 5 were used to fabricate uni-axial dumbbell tensile specimens printed by FDM.
These specimens were designed exploiting the difference in resistivity according to the printing direction, and with dimensions according to ASTM D638, that is to say for the useful part: L = 50mm, W = 10mm , H 300pm. For each material, copies were printed entirely at print directions, measured with respect to the length of the specimen, of 0 °, 45 ° or 90 °. Examples ET1 (0 °), ET2 (90 °) of the manufactured test pieces are illustrated in Figures 6A and 6B.
The traction was exerted through weights hanging vertically at the bottom of the sample, while the top of the sample was attached via a jaw to the frame. The suspended masses were 100 g, 200 g, 500 g and 1 kg. Assuming that the specimens did not have a defect, the stress was concentrated in the useful part, and the stress applied varied between 3 and 33 MPa.
The results of the tests are illustrated in Figures 7A-7C.
FIG. 7A illustrates the evolution of the electrical resistance R of the tensile test pieces of material 2 as a function of the mass M of the weight suspended vertically. The filling rate is 100%, the test is therefore massive.
FIG. 7B illustrates the evolution of the electrical resistance of tensile specimens, under zero load or of 100 g, during a mechanical loading / unloading cycle. The test pieces considered here have a filling ratio of 80%, which means that there is a spacing between the filaments, or streaks, which form the body of the test piece, of which about 20% consists of voids.
FIG. 7B illustrates the evolution in time t of the electrical resistance of the tensile test pieces following the withdrawal of a load of 100 g. Following a sharp increase, the resistance gradually decreases towards the value of rest.
Among the materials studied, Material 3 demonstrated a piezoresistive behavior with threshold effect, when the specimen is printed at 90 °, as shown in Figure 7A. The threshold at which resistance variation occurs can be lowered by decreasing the fill rate when printing by FDM, that is, by modulating the quality of the contact between the streaks. If the fill rate is high, there will be overlap between adjacent streaks. The lower this rate, the thickness of the extrudate being constant, the more the part where there is overlap decreases, and the interface between the adjacent streaks appears holes or pores, where the air is trapped. The porosity rate is therefore an important parameter for controlling the resistance and the piezoresistivity threshold of the printed pattern. After a burn-in cycle (breaking of the weak paths which occurs during the first uses, leading to rapid variations in the electrical properties followed by a stabilization phase), it is noted that the resistance variation appears as soon as 100 g of loading for a bar with a fill rate of 80%, with good repeatability and endurance of at least 10 cycles (see Figure 7B).
The same piezoresistive behavior is observed for the specimens whose streaks are printed at a printing angle of 90 ° with respect to the length of the specimen with the material 5. However, a phenomenon of relaxation of the elastomer matrix following the loading or unloading is observed. For unloading, this phenomenon causes a sharp increase in resistance and a logarithmic decrease towards the resting value (see Figure 7C). This temporal response is thus a significant problem in the reliability of the measurement and its operation to produce a sensor.
The behavioral models and laws established for the piezoresistivity of Material 2, and those established for the resistivity of conductive thermoplastics make it possible to produce sensors that take advantage of these properties.
Figure 8 shows a functional two-dimensional CF2D bending sensor, fully printed by FDM with 3 different thermoplastic materials. This component comprises an insulating thermoplastic SM structure (ABS), four electrically conductive thermoplastic contact electrodes ELC1, ELC2, ELC3, ELC4 (Material 4), and two charged thermoplastic JC1, JC2 core parts having a piezoresistive behavior (Material 2), printed in orthogonal printing directions, forming strain gauges, supported by the mechanical structure SM and whose axes of sensitivity (determined by the printing direction, and thus by the alignment of the constituent filaments) are mutually perpendicular.
In order to avoid the dead time required for coil changes, the sensor was printed using three extrusion heads, one per material. A thin layer of adhesion promoter was deposited at the heterogeneous interfaces.
The central parts of the sensor are respectively a rectangle JC1 of Material 2 printed at 0 ° and a rectangle JC2 of Material 2 printed at 90 °, with a thickness of 600pm. According to the characterization results, only the rectangle whose printing direction is perpendicular to the direction of tensile stress produces a change in electrical resistance. The other block retains the same resistance. The juxtaposition of these two blocks therefore makes it possible to measure a stress along the x-axis or the y-axis, or even a biaxial tensile stress.
When the ABS substrate undergoes a bending stress, its upper face is stressed in tension. This bias is transmitted to the blocks of material 2 by shearing at the interface. The results obtained with this sensor, and demonstrating its functionality, are presented in Table 2 below:
Table 2.
The freedom of form allowed by the process and the availability of conductive and piezoresistive thermoplastics, allows the realization of multidimensional force sensors. FIG. 9F shows a side view of such a sensor in the form of a mini-joystick capable of transducing the force applied on the central connector and its x and y components. This sensor has an insulating mechanical structure comprising an SA annular substrate, PS support pillars and a PTC central platform made of insulating material; piezoresistive elements PZR1 - PZR4 forming suspension bridges made using a printed sacrificial polymer support, for example soluble; contact electrodes EC1 - EC5, central joystick JC and conductive pillars PC1 - PC4 for hooking up the connectors. Figures 9A-9F show these different elements separately.
FIG. 10 illustrates a piezoresistive type acoustic sensor, also produced by a printing method according to the invention. It comprises an insulating mechanical structure SM, two contact electrodes ELC1, ELC2 in conductive thermoplastic, two conductive membranes MCI, MC2, also in conductive thermoplastic, forming the two plates of a capacitor, and strain gauges JC1, JC2, connected each at a respective contact electrode and the conductive membrane MCI. The strain gauges make it possible to measure the deformations of the MCI membrane under the effect of an acoustic wave.
As mentioned above, a design interface is advantageously provided to facilitate the design of printed electrical components. This interface is a computer system (computer, computer network, microprocessor card, etc.) programmed to receive, as input, parameters of an electrical component to be manufactured, such as the desired electrical properties, the position of its grip points. contact, its location within or on the surface of a mechanical structure ... and to provide output, through the application of appropriate algorithms, a print file (typically GCode format) that allows the piloting of the three-dimensional printer.
The flowcharts of FIGS. 11A to 11H illustrate, by way of nonlimiting example, the algorithms making it possible to obtain the printing file for a single electrical component, namely a resistor deposited on an insulating substrate. FIG. 13 illustrates, by way of nonlimiting example, an algorithm making it possible to obtain the print file for an axial strain gauge.
Resistance made by three-dimensional printing can be two-dimensional (i.e. planar, deposited on a surface) or three-dimensional - thus passing through non-coplanar points, which can be imposed. In the case of two-dimensional resistance, one can adopt a linear geometry - for low values of resistance - or in zigzag or serpentine - for higher resistance values. These three different cases require different design algorithms. Thus, as illustrated in FIG. 11A, the user of a design interface according to the invention first defines the thermoplastic material to be used (step EA1), the properties of which are known and stored in the data base of the invention. BDM models, and the resistance value R_user to reach (step EA2), then (EA3) it indicates whether the resistance is three-dimensional or must meet imposed passing points. If so, the design interface will apply a three-dimensional resistor design algorithm, illustrated in Figures 11G and 11H. Otherwise, the system will autonomously determine (EA4), depending on the R_user value, if a linear geometry (for example, if R_user is less than a predefined threshold R_th), in which case it will apply the algorithm illustrated in FIG. 11B, or a zigzag geometry (for example if R_user is greater than or equal to R_th), in which case it will apply the algorithm illustrated in Figures 11C to 11 F. The choice between a linear geometry and zigzag can also take into account other parameters, for example congestion, or be left to the user appreciation.
Whatever the geometry chosen, the applied design algorithm defines the volume of the resistance to be manufactured (EA5). Then (step EA6) a CAD modeller (computer-aided design) generates a triangular mesh of this volume in the form of a triangle, which is typically recorded as an STL stereolithography file. Subsequently (step EA7) is carried out slicing ("slicing" in English) of volume defined by the STL file, which makes it possible to define each slice of material deposited by the extrusion head and to associate therewith a trajectory of moving the print head. The cutting is carried out by a software of a type known per se, depending on the printer, which preferably also makes it possible to parameterize the latter by determining, in particular, the temperature of the edges of the nozzle, the rate or the filling angle, etc. All movements of the nozzle and the machine parameters are recorded in the print file F1, for example in GCode format. The flowchart in Fig. 11B illustrates the linear resistance design algorithm.
To start with, the user supplies as input to the computer system the width Xmax (EB1) and the length Ymax (EB2) of the region, supposedly rectangular, containing the resistance, the position of the point serving as the origin of the reference defined in this region. (EB3), as well as the coordinates of the points of contact A and B of the resistance in this reference (EB4, EB5); these points are usually located on the sides of this rectangular region, in any case it will always be the case later. The computer system then calculates the length L of the segment AB (EB6), then the section S = p L / R_user necessary to obtain the desired resistance (EB7), where p is the resistivity of the material, extracted from the base BDM.
The section S is the product of the width W of the resistance by its thickness H, these two quantities being to be determined. The thickness H of the resistor is an integer multiple of the thickness e of a slice of material deposited by the extrusion head, while the width W is between a minimum value Wmin which is a function of the extrusion head and the permissible width of the resistance, Wmax. We start by assuming H = e, and we calculate the corresponding width: W = S / H. If a value greater than the maximum permissible width Wmax is found, H is incremented by the value e, and so on. Although this is not shown in the figure, it is possible that even with H = e we get W <Wmax. This means that R_user is too high; it is therefore necessary either to increase the length L by moving one of the contact points (or both), to change the material, or to use a zigzag geometry. In principle, this situation should not occur if the linear geometry has been automatically selected by the algorithm of Figure 11 A. These operations constitute step EB8.
Then (EB9) we check that the value R = RHO L / S is equal to R_user. If this is not the case, the user has the possibility of changing the material used, in which case the algorithm resumes from step EB7; if it does not wish to do so, the computer system moves one of the points A and B (or both) to change the length L until R = R_user (EB10). The algorithm for designing a zigzag resistance is illustrated in FIGS. 11C-11F.
As in the case of the linear resistance, the user supplies as input to the computer system the width Wmax (EC1) and the length Ymax (EC2) of the region, supposed rectangular, containing the resistance, as well as the coordinates of the origin of the marker (EC3). It also provides the width W and the thickness H of the resistive "wire" forming the zigzag pattern (EC4, EC5), which allows the system to calculate the length of this wire: L = R_user-WH / RHO (EC6 ). Then, the user inputs the coordinates of the contact points A and B of the resistor (EC7, EC8), which allows the system to determine if these points are on opposite sides of the rectangular region, on the same side or on adjacent (and orthogonal) sides. To these three cases correspond three different algorithms, which are illustrated by Figures 11D, 11E and 11F, respectively.
When the point B is in front of A (FIG. 11D), the path of the conducting wire is determined starting from the point A (step EDI), by moving with a predefined pitch of length DELTA_min (small compared to the dimensions of the rectangular region containing the resistance, but greater than the width W of the wire) to the side where B is and perpendicular to the side where A (ED2) is. Then, one moves parallel to the side of A, in the direction of the vertex of this side furthest away from A, stopping at a distance DELTA_min from the edge of the rectangular region, to keep a margin of safety (ED3). Then we move again the length DELTA_min to the side where B is and perpendicular to the side where is A (ED4). The distance traveled from the start is then stored, and stored in a variable dist_p (ED5). At this point, the minimum distance remaining in a straight line is calculated to reach B (ED6). If the sum of dist_p and this minimum distance exceeds L (calculated in step EC6), it means that it is not possible to achieve a zigzag resistance of the desired value; you have to change the material. Otherwise, the system determines whether points A and B are facing in the x direction or the y direction. In the following, only the first case is considered (steps ED7, ED8, ED9); the operations performed in the opposite case are quite similar (steps ED7 ', ED8', ED9 ').
The zigzag resistance consists essentially of a number of patterns (meanders) formed of a short line, of length d, in the direction perpendicular to the sides which bear A and B (direction x, in the example considered here) and a long line, of length D, in the perpendicular direction, plus two end segments - of a length dist_p on the side of A and a length to be calculated on the side of B. When of step ED7, we try to solve the system: nb_p (d + D) <L-dist_p nb_pd <Xmax-2DELTA_min by making an assumption on the value of D (because we have three unknowns and only two inequalities). The initial hypothesis is: D = Ymax-2DELTA_min.
During step ED8 it is checked whether the solution found verifies the condition d> DELTA_min. If this is not the case, the value of D is decremented by DELTA_min (ED9), and the step ED7 is executed again.
The final section consists of the shortest path that connects the terminal end of the last pattern to B (ED10). In general, to obtain a total length equal to L, it will be necessary to adjust the length of the last long line (penultimate section of the resistance, which will not necessarily be equal to D. To build the volume of the resistance from the course thus determined it remains only to apply the width of the wire W (ED11).
When the point B is on the same side as A (FIG. 11E), the path of the conducting wire is determined starting from the point A (step EE1), moving with a predefined pitch length DELTA_min (small compared to dimensions of the rectangular region containing the resistance, but greater than the width W of the wire) in the direction perpendicular to the side where the points A and B (EE2) are located. Then (EE3) one moves parallel to this side and the opposite of B, until leaving only a safety distance DELTA_min of the side perpendicular to that of departure. Then we move further the distance DELTA_min in the direction of the first displacement, that is to say perpendicular to the bearing side A and B (EE4). The following steps - EE7, EE8 to EE10, EE8 'to EE10', EE11, EE12- are identical to the steps ED6, ED7 to ED9 and ED7 'to ED9', ED10 and ED11 described above.
When the point B is on an adjacent side, and therefore perpendicular to the one bearing the point A (FIG. 11F), the path of the conducting wire is determined starting from the point A (step EF1), moving from one not predefined length DELTA_min (small relative to the dimensions of the rectangular region containing the resistance, but greater than the width W of the wire) in the direction perpendicular to the side where is the point A (EF2), then moving parallel to this side and towards the side opposite to the point B, up to a safety distance DELTA_min of the latter (EFS). Then we move parallel to this side until a safety distance DELTA_min left opposite to that of A (EF4). The sequence (steps EFS - EF11) is identical to the previous cases.
The last case, that of a three-dimensional resistance, is the subject of FIGS. 11G and 11F1. First, the user must define a volume in which the resistance is defined, the starting points (A) and of arrival (B), as well as intermediate points of passage (EG1). It is important to note that in the three-dimensional printing technique FDM, a volume is in the form of a stack of layers of predefined thickness. The position of a point can therefore be defined by identifying the layer to which it belongs and by giving its two-dimensional coordinates inside the latter. To define the path of the resistance, we start from the point A (EG2), we identify the projection pt_s of the next point of passage on the layer containing A - or, more generally, the current point pt_c (EG3), and we determine the distance to reach this projection (EG4), as well as the difference in height between the layer of the current point and that of the next point (EG5). The distance L_cs between the points pt_c and pt_s is the sum of the distance in the plane, determined in step EG3, and the difference in height (EG6); this distance is stored. Then we go to the next point, and so on, until we reach the point B. We then calculate L_calc, the sum of the lengths L_cs (EG7). Knowledge of the RFIO resistivity of the material then makes it possible to calculate the section S of the resistive pattern (EG8); from the knowledge of this section, the thickness and the width are calculated as explained with reference to the linear case of FIG. 11B (step EG9). We then check whether the resistance thus obtained, R_calc, has the desired value R_user (EG10). If R_calc, <R_user, and if the user does not wish to change material, it is necessary to replace at least some of the straight segments connecting two successive crossing points within the same layer by zigzags. To do this, the length of each intermediate segment is expressed as a percentage of L_calc (EG11), and it is assigned a resistance to reach which corresponds to the same percentage of R_user (EG 12). Then, as in the case of FIG. 11D, the steps EG13-EG16 correspond to the steps ED2-ED5, the steps EG17, EG18, EG19 to the steps ED7 to ED9 and ED7 'to ED9' and the steps ED20, ED21 to the steps EDIOetEDII.
Another example of an electrical component that can be manufactured by three-dimensional printing according to the invention, and whose manufacture can be facilitated by the use of a programming interface, is a uni-axial deformation (or stress) type gauge. piezoresistive.
The relative resistance variation ΔΒ of a piezoresistive element having a resting resistance Ro is: AR / Ro = k-ε where k is the piezoresistive sensitivity and the strain, with s = AL / Lo, where AL is the variation in length and L the length at rest of the element. When designing a sensor, the initial length (and therefore the resistance) is a function of the fill rate of the sensor volume.
Like any element made by three-dimensional FDM printing, a piezoresistive sensor is formed by filaments, or bands; this structure is illustrated in Figure 12A. The sensor works by bringing the bands closer together or away from one another under the effect of a constraint. Indeed, if we bring the bands together, we create microcontacts that make the resistance fall. Conversely, if the bands are moved away, the current intensity that circulates is lower. It is understood that the sensitivity to the deformations is much greater for the deformations parallel to the small axis of filling, indicated by d in FIG. 12A, than for the deformations parallel to the major axis D.
If the filling ratio is 100%, which means that the bands are perfectly joined and in contact with each other over their entire length, the resistance variations will not be measurable. It will be the same in the case of a fill rate too low. Empirically it has been found that fill rates of 40 to 70% give quite acceptable results for most of the considered materials.
A comparison of Figures 12A and 12B shows that a 40% fill (Figure 12A) implies a shorter filament length than with a 70% fill (Figure 12B). In a sensor with a filling rate of the order of 70%, the conductivity depends significantly on the percolation of the current, that is to say the presence of conducting microchips connecting the meanders with each other. These micro-paths can be easily interrupted by small deformations of the structure, so that a sensor having a high filling rate - but not too much, for example of the order of 70% - will be able to detect moreover small deformations as a sensor with a fill rate of only 40%. However, such a sensor will also be more sensitive to manufacturing defects, for example, short circuits caused by a dirty extrusion head. It should be noted that in the examples of FIGS. 12A and 12B, several layers of resistive material have been deposited, with perpendicular deposition angles, which leads to an isotropic piezoresistive sensitivity. The flow chart of Figure 13 illustrates the design steps of a piezoresistive stress sensor. First (EPI step) The user realizes the desired shape on a CAD software and imports it in STL format (mesh with triangles of the volume). Then, in order to be sure that the sensor will be sufficiently flexible and resistant, its height H is fixed at a predefined value, for example three times the printing thickness e (EP2).
The last layer will serve as a sensor. The other layers may for example be made of insulating material with a fixed filling, generally between 60 and 100% and for example equal to 70%, so as to leave more or less flexibility to the sensor (EPS). Alternatively, conductive or resistive layers having another function may also be present below that serving as a strain sensor.
On the sensor layer, the user defines the position of the electrodes (EP4) and chooses the resistive material from among a proposed list (EPS) and the desired sensitivity - also from a list of possible values (EP6) . The filling is generally in rectilinear mode (EP7), with a main filling direction perpendicular to the axis defined by the two electrodes (EPS).
A table, built during a preliminary calibration step and stored in a memory of the computer executing the process, makes it possible to find the filling ratio making it possible to reach the desired sensitivity, given the chosen material and the desired geometry. (EP9).
The print speed (EP10) can be determined according to the fill rate also determined, typically by means of another correspondence table constructed by calibration. For example, a low print speed can be chosen for large fill rates to avoid short circuits that may occur when filament residues remain stuck to the nozzle and unintentionally peel off during printing. . For lower filling rates, the spacing of the passage lines, and therefore the absence of material in the passage of the nozzle, is generally sufficient to prevent a residue from sticking; it is therefore possible to print at a higher speed, for example 60 mm / s. Alternatively, a relatively low print speed (e.g., 30 mm / sec) may be used regardless of the fill rate. The invention has been described with reference to mechatronic systems comprising piezoresistive sensors. However, the invention is not limited to this case. It is indeed possible, in particular by choosing in a timely manner the fillers contained in the thermoplastic matrix, to print by LFL materials sensitive to temperature, light, chemical agents, etc. in order to obtain thermal, optical and chemical sensors, respectively. Moreover, from conductive and insulating materials, it is possible to manufacture capacitive or inductive sensors, and antennas that can be used for transmitting and / or receiving. It is also possible to produce thermal actuators, electrostatic, magnetic (using ferromagnetic loads) or even piezoelectric. It is also possible to print OLEDs (Organic Light Emitting Diodes) and / or organic photovoltaic cells.
In some cases, the invention will allow the realization of a fully printed mechatronics system. In other cases, only a part of the electrical / electronic subassembly of such a system will be printed (sensors, conductive lines, actuators ...) while other components - for example integrated circuits performing electronic functions complex, may be reported. Even in the latter case, the invention makes it possible to reduce the number of assembly operations, and thus to make the manufacture of the mechatronic system faster and less expensive. References [Coiai, 2015] Serena Coiai and. Al, "Nanocomposites Based on Thermoplastic Polymers and Functional Nanofiller for Sensor Applications," Materials, vol. 8, pages 3377-3427, 2015.
[Deng, 2014] Hua Deng et al., "Progress on the morphological control of conductive network in conductive polymer composites and the use of electroactive multifunctional materials," Progress in Polymer Science, Vol 39, pp. 627-655, 2014.
[Muth, 2014] Joseph T. Muth et al., "Embedded 3D Printing of Strain Sensors Within Highly Stretchable Elastomers," Advanced Materials, Vol 26, pp. 6307-6312, 2014.
[Sitthi-Amorn, 2015] Pitchaya Sitthi-Amorn et al. "MultiFab: A Machine Vision Assisted Platform for Multi-material 3D Printing" ACM Transactions on Graphics (SIGGRAPH 2015).
[Rossiter, 2009] Jonathan Rossiter et al. "Printing 3D dielectric elastomer actuators for soft robotics" Proc, of SPIE Vol. 7287, 72870H (2009).
[Leigh, 2012] Simon J. Leigh et al. "A Simple, Low-Cost Conductive Composite Material for 3D Printing of Electronic Sensors", PLOS ONE, Vol. 7, Issuer 11, e49365 (Nov. 2012).
[O'Brien, 2015] Jonathan O'Brien et al. "Miniaturization of Microwave Components and Antennas Using 3D Manufacturing" 9th European Conference on Antennas and Propagation (EuCAP), Lisbon, May 7, 2015.

权利要求:
Claims (15)
[1" id="c-fr-0001]
A method of manufacturing a mechatronic system comprising: a step of manufacturing a mechanical structure (SM) by three-dimensional printing by deposition of molten wire of at least a first electrically insulating first material (M1); and a step of manufacturing at least one electrical component (CE) in contact with at least one element of said mechanical structure and integral therewith; characterized in that said step of manufacturing at least one electrical component is implemented by three-dimensional printing by deposition of molten wire of at least a second material (M2), conductive or resistive, directly in contact with said element of the structure mechanical.
[2" id="c-fr-0002]
2. The method of claim 1 wherein said or one said electrical component is a transducer (JC1, JC2).
[3" id="c-fr-0003]
3. The method of claim 2 wherein said transducer (JC1, JC2) is a piezoresistive sensor.
[4" id="c-fr-0004]
4. Method according to one of the preceding claims wherein said second material, conductive or resistive, comprises conductive fillers (CM2) dispersed in a thermoplastic insulating matrix (MM2).
[5" id="c-fr-0005]
5. Method according to one of the preceding claims also comprising a local annealing step implemented during or after the deposition of a layer of the first or second material, in correspondence of said deposit.
[6" id="c-fr-0006]
6. Method according to one of the preceding claims also comprising a step of deposition of an adhesion promoter agent (APA) on a surface of the mechatronic system during manufacture before the deposition, above said surface, of a layer of a different material.
[7" id="c-fr-0007]
7. Method according to one of the preceding claims comprising the use of at least two separate extrusion heads (TE1, TE2) for depositing the first and second material.
[8" id="c-fr-0008]
8. Method according to one of the preceding claims also comprising a step of generating a print file (Fl) for producing at least one said electrical component, said step being implemented by computer and comprising: a sub-step of providing to said computer data indicative of a position of one or more contact points, a spatial region where said component is to be manufactured and at least one electrical property of said component; a sub-step of calculating a geometry of said component by applying to said data a predefined mathematical model; and a sub-step of generating said print file making it possible to produce said geometry by three-dimensional printing by deposition of said or said at least one second conductive or resistive material.
[9" id="c-fr-0009]
A mechatronic system (CF2D) comprising an electrically insulating mechanical structure (SM) and at least one electrical component (CE) arranged in contact with at least one element of said mechanical structure and integral with it, characterized in that the mechanical structure and the electrical component are made in one piece by three-dimensional printing of at least one electrically insulating first material (M1) forming said mechanical structure and at least one second conductive or resistive material (M2) forming said electrical component.
[10" id="c-fr-0010]
10. Apparatus for implementing a method according to one of claims 1 to 8 comprising a three-dimensional printer (I3D) of the type of deposition of molten wire having at least two separate extrusion heads (TE1, TE2), independently activatable and adapted to deposit two different materials, said extrusion heads being arranged side by side with the same extrusion direction and being carried by the same printing carriage (C1) ensuring their simultaneous movement, the head of printing also comprising a mechanism (ATV) for moving an inactive extrusion head in a direction opposite to said extrusion direction when said one or said other extrusion head is active.
[11" id="c-fr-0011]
The apparatus of claim 10 wherein said print carriage is equipped with a capacitive sensor (CC) configured to measure its distance from a printing surface.
[12" id="c-fr-0012]
12. Apparatus according to claim 11 also comprising a printing plate (PLI) above which moves said printing carriage, said plate being equipped with metal electrodes (EET1, EET2) capable of being detected by said sensor capacitive, whereby said sensor allows a calibration of the position of the print carriage relative to the plate.
[13" id="c-fr-0013]
13. Apparatus according to one of claims 10 to 12, equipped with: a camera (CE) configured to acquire an image of a layer of material deposited by the three-dimensional printer; an image processing system (STI) configured to compare said image with a three-dimensional model stored in a computer memory and derive a position error from said print carriage; and a computer control system (SIP) of said carriage configured to correct said position error during the deposition of a successive layer of material.
[14" id="c-fr-0014]
14. Apparatus according to one of claims 10 to 13 also comprising a computer system (SGF) for generating a print file (Fl) for driving said three-dimensional printer so as to manufacture an electrical component, said computer system being configured to: input data indicative of a position of one or more contact points, a spatial region where said component is to be manufactured and at least one electrical property of said component; computing a geometry of said component by applying to said data a predefined mathematical model; and generating a print file for making said geometry by three-dimensional printing by depositing molten wire of at least one conductive or resistive material.
[15" id="c-fr-0015]
15. Apparatus according to one of claims 10 to 14 also comprising a generator (SRR) of an electromagnetic radiation beam, said generator being configured to produce a local heating of a material deposited on a printing surface of said printer three-dimensional.

类似技术:

---

公开号 | 公开日 | 专利标题

---

FR3052698B1|2019-08-09|METHOD AND APPARATUS FOR MANUFACTURING A MECATRONIC SYSTEM BY THREE-DIMENSIONAL PRINTING

---

US9981314B2|2018-05-29|Direct writing for additive manufacturing systems

---

Riul et al.2002|Artificial taste sensor: efficient combination of sensors made from Langmuir− Blodgett films of conducting polymers and a ruthenium complex and self-assembled films of an azobenzene-containing polymer

---

Beedasy et al.2020|Printed electronics as prepared by inkjet printing

---

Vaithilingam et al.2017|Combined inkjet printing and infrared sintering of silver nanoparticles using a swathe-by-swathe and layer-by-layer approach for 3-dimensional structures

---

FR3024303A1|2016-01-29|IMPROVED METHOD FOR MAKING A TRIBOELECTRIC GENERATOR WITH ROUGH DIELECTRIC POLYMER

---

Chang et al.2015|A facile method for integrating direct-write devices into three-dimensional printed parts

---

CN104407409B|2017-01-11|3D printing method for dynamic liquid crystal grating

---

Nesser et al.2018|Towards wireless highly sensitive capacitive strain sensors based on gold colloidal nanoparticles

---

Rajan et al.2017|WORM and bipolar inkjet printed resistive switching devices based on silver nanocomposites

---

WO2018074993A1|2018-04-26|Hybrid fusion system

---

Borghetti et al.2019|Printed strain gauge on 3D and low-melting point plastic surface by aerosol jet printing and photonic curing

---

Zhang et al.2021|Reviews on Machine Learning Approaches for Process Optimization in Noncontact Direct Ink Writing

---

Milano et al.2020|Mapping time-dependent conductivity of metallic nanowire networks by electrical resistance tomography toward transparent conductive materials

---

Zhang et al.2018|Heterogeneous integration of three-primary-color photoluminescent nanoparticle arrays with defined interfaces

---

Santaniello et al.2020|Additive nano-manufacturing of 3D printed electronics using supersonic cluster beam deposition

---

Jahangir et al.2020|On Self-Limiting Rotation and Diffusion Mechanisms during Sintering of Silver Nanowires

---

Kim et al.2020|A fluoropolymer-coated nanometer-thick Cu mesh film for a robust and hydrophobic transparent heater

---

Liu et al.2021|An Integrative 3D printing method for rapid additive manufacturing of a capacitive force sensor

---

Bridonneau et al.2020|Morphological control of linear particle deposits from the drying of inkjet-printed rivulets

---

Thetpraphi et al.2021|3D-printed electroactive polymer force-actuator for large and high precise optical mirror applications

---

Hines et al.2021|Considerations of aerosol-jet printing for the fabrication of printed hybrid electronic circuits

---

Sopoušek et al.2021|Thick nanoporous matrices of polystyrene nanoparticles and their potential for electrochemical biosensing

---

Chang2017|Process monitoring, modeling, and quality assessment for printed electronics with aerosol jet printing technology

---

FR3087698A1|2020-05-01|MANUFACTURE OF PIEZOELECTRIC COMPONENTS AND DEVICES WITH A THREE-DIMENSIONAL PRINTING PROCESS

同族专利:

---

公开号 | 公开日

---

ES2785153T3|2020-10-06|

---

JP2019518633A|2019-07-04|

---

TW201803723A|2018-02-01|

---

US20190217529A1|2019-07-18|

---

BR112018075748A2|2019-03-26|

---

SG11201811222VA|2019-01-30|

---

AU2017286162A1|2019-01-31|

---

WO2017215971A1|2017-12-21|

---

IL263692D0|2019-01-31|

---

EP3471947A1|2019-04-24|

---

CA3027596A1|2017-12-21|

---

EP3471947B1|2020-02-19|

---

KR20190019076A|2019-02-26|

---

FR3052698B1|2019-08-09|

---

CN109562562A|2019-04-02|

---

US10807301B2|2020-10-20|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

US20100021580A1|2008-07-25|2010-01-28|Swanson William J|Pantograph assembly for digital manufacturing system|

---

WO2014209994A2|2013-06-24|2014-12-31|President And Fellows Of Harvard College|Printed three-dimensional functional part and method of making|

---

US20150173203A1|2013-12-12|2015-06-18|Xyzprinting, Inc.|Three-dimensional printing apparatus|

---

US20160151978A1|2014-11-12|2016-06-02|Etron Technology, Inc.|Three-dimensional printer with adjustment function and operation method thereof|CN110978511A|2019-12-23|2020-04-10|芜湖市爱三迪电子科技有限公司|Line footpath controlling means of 3D printing consumables|

---

WO2020222762A1|2019-04-29|2020-11-05|Hewlett-Packard Development Company, L.P.|Three-dimensional printing|

---

WO2021069821A1|2019-10-09|2021-04-15|Commissariat A L'energie Atomique Et Aux Energies Alternatives|Composite material for 3d printing, and 3d printing method|AU2012214512B2|2011-02-07|2015-11-19|Ion Geophysical Corporation|Method and apparatus for sensing underwater signals|

---

US20130142942A1|2011-11-17|2013-06-06|Abbott Diabetes Care Inc.|Methods of Making a Reference Electrode for an Electrochemical Sensor|

---

US9149988B2|2013-03-22|2015-10-06|Markforged, Inc.|Three dimensional printing|

---

US9499896B2|2013-09-18|2016-11-22|Neumodx Molecular, Inc.|Thermocycling system, composition, and microfabrication method|

---

KR102327600B1|2013-12-26|2021-11-16|텍사스 테크 유니버시티 시스템|Microwave-induced localized heating of cnt filled polymer composites for enhanced inter-bead diffusive bonding of fused filament fabricated parts|

---

US20150201500A1|2014-01-12|2015-07-16|Zohar SHINAR|System, device, and method of three-dimensional printing|

---

FR3029838A1|2014-12-11|2016-06-17|Centre Nat Rech Scient|METHOD FOR ADDITIVE MANUFACTURING OF A 3D MECATRONIC OBJECT|

---

WO2016171623A1|2015-04-20|2016-10-27|Agency For Science, Technology And Research|Conductive polymer composite as plastic solder|

---

US10808081B2|2015-09-04|2020-10-20|Sabic Global Technologies B.V.|Powder compositions comprising thermoplastic particles and flow promoter particles, method of preparing articles and coatings from the powder compositions, and articles prepared therefrom|WO2019198039A2|2018-04-14|2019-10-17|Lohia Corp Limited|An integrated intelligent system and a method for estimating actual thickness of extruded films|

---

EP3576503A1|2018-05-31|2019-12-04|Airbus Operations, S.L.|Implementation process of a humidity control sensor for 3d-printing of a smart tool and process for manufacturing a humidity control sensor via 3d-printing technology|

---

CZ308757B6|2020-03-03|2021-04-28|Západočeská Univerzita V Plzni|Resistor manufacturing method for power applications|

---

DE102020106559A1|2020-03-11|2021-09-16|Ntt New Textile Technologies Gmbh|Textile carrier material|

---

LU101897B1|2020-06-27|2022-01-10|BigRep GmbH|3D-printing system and method|

法律状态:
2017-05-30| PLFP| Fee payment|Year of fee payment: 2 |

---

2017-12-22| PLSC| Search report ready|Effective date: 20171222 |

---

2018-05-25| PLFP| Fee payment|Year of fee payment: 3 |

---

2019-05-22| PLFP| Fee payment|Year of fee payment: 4 |

---

2021-03-12| ST| Notification of lapse|Effective date: 20210206 |

优先权:

---

申请号 | 申请日 | 专利标题

---

FR1655552A|FR3052698B1|2016-06-15|2016-06-15|METHOD AND APPARATUS FOR MANUFACTURING A MECATRONIC SYSTEM BY THREE-DIMENSIONAL PRINTING|

---

FR1655552|2016-06-15|FR1655552A| FR3052698B1|2016-06-15|2016-06-15|METHOD AND APPARATUS FOR MANUFACTURING A MECATRONIC SYSTEM BY THREE-DIMENSIONAL PRINTING|

---

JP2018565811A| JP2019518633A|2016-06-15|2017-06-05|Method and apparatus for manufacturing mechatronic system by three-dimensional printing|

---

ES17727246T| ES2785153T3|2016-06-15|2017-06-05|Procedure and apparatus for the manufacture of a mechatronic system by three-dimensional printing|

---

PCT/EP2017/063606| WO2017215971A1|2016-06-15|2017-06-05|Process and apparatus for manufacturing a mecatronic system by three-dimensional printing|

---

US16/307,132| US10807301B2|2016-06-15|2017-06-05|Method and apparatus for manufacturing a mechatronic system by three-dimensional printing|

---

CN201780049637.1A| CN109562562A|2016-06-15|2017-06-05|Method and apparatus for manufacturing Mechatronic Systems by 3 D-printing|

---

EP17727246.5A| EP3471947B1|2016-06-15|2017-06-05|Process and apparatus for manufacturing a mecatronic system by three-dimensional printing|

---

SG11201811222VA| SG11201811222VA|2016-06-15|2017-06-05|Method and apparatus for manufacturing a mechatronic system by three-dimensional printing|

---

CA3027596A| CA3027596A1|2016-06-15|2017-06-05|Procede et appareil pour la fabrication d'un systeme mecatronique par impression tridimensionnelle|

---

AU2017286162A| AU2017286162A1|2016-06-15|2017-06-05|Process and apparatus for manufacturing a mecatronic system by three-dimensional printing|

---

KR1020187036455A| KR20190019076A|2016-06-15|2017-06-05|Process and apparatus for manufacturing mechatronic systems by 3D printing|

---

BR112018075748-9A| BR112018075748A2|2016-06-15|2017-06-05|Method and apparatus for the manufacture of a three-dimensional printing mechatronic system|

---

TW106119679A| TW201803723A|2016-06-15|2017-06-13|Method and apparatus for manufacturing a mechatronic system by three-dimensional printing|

---

IL263692A| IL263692D0|2016-06-15|2018-12-13|Process and apparatus for manufacturing a mecatronic system by three-dimensional printing|