950829 Plastic Windshield Wiper Arm Géraldine Cervantes and Bruno Lisiecki Valeo ABSTRACT To reduce the weight of the front windshield system, a fiber reinforced plastic design for the wiper arm and blade is being investigated. The use of suitable thermoplastics reinforced with short fiberglass allows the elaboration of good stiffness parts by injection moulding. Tests were carried out to show that stiffness depends on the fiberglass concentration and orientation. Likewise the temperature leads a decreasing of mechanical properties. Processing conditions of plastics have significant influence and aim to increase their performances, effectively a distribution in the orientation of fiberglass can be observed in the thermoplastic matrix. Various experimental means are used to characterise the plastic materials. INTRODUCTION Wiper system development is based on four targets: reliability, style, weight reduction and cost. The goals are main parameters in the choice of materials used in the wiper system. A rear windshield wiper arm with reinforced polyester has been created. It is planned to use a fiber reinforced thermoplastic for a front windshield wiper arm. Important researches are made to develop new products. Traditional materials will be replaced by some plastic materials even if environmental conditions are drastic i.e. high mechanical stresses, chemical and physical constraints. Only some specific plastics can comply with these constraints and present such advantages as corrosion resistance and style facilities which is not the case of metallic structures. Suitable plastic materials will allow to reach these new specifications and new characteristics, effectively fiberglass reinforced plastics present good stiffness properties, thermal and bad weather resistance. Design engineering of windshield wiper arm is a hierarchical process that relies on computer aided methods for engineering analysis and simulation provided that mechanical properties used in the calculation are accurate. Today the motor world shows a strong tendency to improve the performances of cars and to limit weight parts at the same time. In this objective and because the style is not focused any more only on the body of cars Valeo has the challenge to design a front wiper system in plastic materials. Designing a front plastic wiper arm assembly requires a perfect skill mastery in terms of parts dimensioning, mechanical constraints and choice of a high tensile reinforced material. Our investigation is composed of three parts. First it was necessary to seek out the material which answers to all the specifications: standard tests were carried out on different materials to determine mechanical properties. In a second time the strength of material is planned to be investigated. A study will enable to determinate and to locate constraints on different shapes. It is a real necessity to optimise the shapes in order to improve mechanical properties. Finally a rheological study will allow to optimise the injection moulding process, find the weldlines (1) and determinate the best locations for the injection points to fill the parts. A study has been started to analyse the rheological behavior during the injection process thanks to an instrumented mould. In this paper the choice of the material, the kind of reinforcement and fiberglass orientation are discussed. THE WIPER ARM AND BLADE The wiper arm assembly is a technical part which support many stresses. The wiper arm and blade assembly is part of the external component of the 792 vehicle. The part that actually makes contact with the surface of the glass and does the wiping is the rubber portion of the wiper blade. The blade repeats a smooth side-to-side (or up and down) movement; the pressure distribution relative to the glass curve must be as even as possible. The wiper blade needs to be kept under pressure. The rubber blade stays on a curved windshield thanks to a spring located under the system. The arm must be rigid enough in static conditions (no creep is accepted) and also under dynamic conditions to avoid mechanical fatigue. To ensure a good pressure distribution, all the wiper arm component must be made of materials with suitable tribological and dimensional properties. Since wiper blades are outside the car, bad weather resistance and windshield liquid resistance must be excellent. Working specifications impose a wiping quality under high speed of the car (80% of the maximum speed limit of the car) which is difficult to reach because of aerodynamic constraints. Kinetic energy produced by wiper engine is veiy high and constraints are concentrated in specific part location. Other parameters of the working specifications impose a good high temperature stability and impact resistance. WHY USE PLASTIC MATERIALS As noticed in the introduction, weight reduction is a main line research for the parts design. Plastic density is in a perfect agreement with this principle. The present front wiper system is made with metal for the arm and metal with plastic parts for the blade. The weight of the complete system is about 0,450 kg (1 pound). We expected 20% weight reduction by using plastics for both arm and blade. To reinforce a plastic material one can use different kind of fibers: glass (long or short), carbon or kevlar. Concerning the matrix, thermoplastic or thermosetting plastics are available. Because production cycles must be veiy competitive in the industrial world and the cost as low as possible, the use of short fiberglass reinforced thermoplastics is an obligation. Plastics provide a great design versatility (hot plastic fluidity allows to fill complex shapes in mould). Also the melt temperature of plastic is lower than for metal so that the energy required is lower and cycles are shorter. Specific strength of plastic makes easier the design of one single part without oversizing the system (time assembly is reduced). Also one single part could favour the use of one material only and should facilitate the recycling. By choosing a UV stabilised grade (black for our application) one avoid painting and anti-corrosion process. ADVANTAGES OF PLASTIC MATERIALS COMPARED TO METALS Polymer density is very interesting in terms of weight reduction, but it is necessary to estimate the rigidity of plastics by studying their specific moduli. Specific properties are given in table 1. There is a 4.5 ratio between steel and plastic so that using plastic will impose an appropriate dimensioning. A very simple calculation can illustrate the above: Suppose a section sized with steel (like the arm section is nowadays), the geometry is a (9*3) mm rectangle (= 0.35*0.12 inches). If you want a section in plastic which is as rigid as the steel one, you need a (6*17) mm rectangle (= 0.24*0.67 inches). We obtain the same rigidity with both materials but: - plastic allows a 15 % reduction of weight, - steel allows a lower dimension. Table 1: Mechanical properties of technical materials (2). Materials Density Modulus Specific Ultimate Specific kg/m3 (GPa) modulus Strength strength Aluminium 2700 71 26 270 100 Steel 7860 210 27 460 59 Polyamide 66 U4Õ 3 Z6 80 70 PA66 + 30% gf 1380 8 5Ü 160 U6 Polycarbonate 1240 2.3 1.9 60 48 Compared to metallic structures, pla a longer development and a more e investment. This cost decreases fast production run. Valeo plans to pro radius systems with plastic materi radius generating high mechanical con FIBERGLASS REINFORCEMENT FOR THERMOPLASTICS Stiffness and lightness are the major parameters to investigate when one wants to replace metallic materials with plastics. How is it possible to improve plastics rigidity ? First by choosing an appropriate material. Tensile tests and bend tests on different kind of materials were carried out. Another good way is to optimise the anisotropy of the material. Metal have an isotropic behavior because they are polycrystalline. Long fibers reinforced thermoplastics are totally anisotropic. Using a short glass fibers reinforced plastic gives the possibility to generate located anisotropy, thanks to the orientation of fibers. 793 Improvement of the rigidity and high temperature properties of the polymer can be achieved by working on its cristallinity but thermal conditions of moulding process will limit this possibility. Fiber reinforcement is more easily exploitable. Glass fiber gives rigidity and toughness while thermoplastic matrix gives the impact and heat resistance. Bare fibers are very brittle so one need to optimise their quantity to use their properties in the best way. EFFECT OF FIBERS ON MODULUS - Tensile tests with a dynamometer (DY 25 from Adamel Lhomargy) instrumented with an extensometer (EX 10 from Adamel Lhomargy) were carried out to show the effect of fiberglass rate on the Young modulus at room temperature (NF standard n° NF T 57101). Materials are produced by a great chemical company. Results are given in table 2 and studied in figure 1. Table 2: Experimental Young modulus of technical plastics (3). i ' - - - Materials Young modulus(MPa) PBT 20 % glass fiber PBT 30% glass fiber PBT 50 % glass fiber PA6/6T 25% glass fiber PA 6/6T 35 % glass fiber PA 6/6T 50 % glass fiber | 14930 Young modulus in function of fiber rate T ensile tests at room temperature 20 1 'cř CU 2 15 - 1 J 10 - ♦Polyester (PBT) I ® AA rom a tic polyamide PA 6/6T I I s,A 1 oL - I 0 10 20 30 40 Glass fiber volum ic rate (%) Figure 1: Young modulus evolution in function of glass fiber rate. EFFECT OF FIBER ORIENTATION - Short glass fibers are special reinforcement because they bring by their own nature a stiffener which can generate some anisotropy. We highlight anisotropy with two kind of tests: tensile tests on micro tensile specimens and ultrasonic measurements on small specimens. Dynamometer measurements - Fiber orientation is responsible for anisotropy especially if the mould filling shows a jetting phenomenon (3). Jetting is very frequent with reinforced materials (because the extruded part swelling is limited). We underline jetting with Young modulus measurements on fatigue standard test specimens. Normalised (ASTM standard n° D 1708 ) tension test specimens A, B and C were cut in fatigue standard specimens moulded with PBT , as shown in figure 2, and were tested on a dynamometer. Results are given in table 3. I m r^A ! ^ MoiteiÉfedwintÉkfiilfli kdbi B c Esļļga lUfeaevedimM M^ikL PBT3I* »h» a>m Figure 2: Description of the position of micro tensile specimens. Table 3: Effect of fiber orientation on rigidity (3). Position on samples Rigidity Breaking load parameter (MPa) Material :PBT 30% gf Position A Position B Position C Material: PBT 50% gf Position A Position B Position C N.B.: The rigidity p displacement/specime Because of jetting, f directions according Results show that the quite the same: fiber 794 direction of the micro tensile specimens. By using a dynamometer we only determinate an overall Young modulus and sample geometry is imposed by the standard. Ultrasonic measurements - Orientation was also proved thanks to Ultrasonic modulus measurements. This technique (4), easy and fast, consists in measuring longitudinal ultrasonic wave propagation velocity. A transducer generates ultrasonic wave on one side of the part, waves are collected on the other side. The time needed by the waves to cross the material length allows the calculation of modulus. Wave frequency is high so that a dynamic modulus is measured. Samples can be very small, provided that the section is constant. Thus it is possible to cut samples anywhere in the injected part in order to determine local modulus as shown in figure 3. j r~i ^2^-- B Fatigue ^ecimen fett Material: PBT 3044 glass fiber Finire 3: Description of the specimens cut for the ultrasonic analysis. Ultrasonic results are given in table 4. In case of random orientation of fibers we should have found the same modulus for samples A and B. Despite the last scatter associated with this technique our results show a different modulus for A and B. From this, we can conclude that fibers are oriented in a specific direction. Table 4 : Modulus measurements with ultrasonic method. I Material Sample used Sample Sample : PBT B A 30% Glass Material used : PBT 30% Glass Sample A 7860 - 8760 Sample B 5650 - 5910 As shown in figure 4, the analysis of modulus shows fibers are nearly all oriented longitudinally alon flow because the modulus of A is a little bit lower than the longitudinal Halpin-Tsaï modulus, while B modulus is a little bit higher than the transversal Halpin-Tsaï modulus (see next paragraph for the definition of the Halpin-Tsaï modulus). Figure 4: Halpin-Tsaï theoretical modulus versus ultrasonic experimental modulus. Theoretical models - Halpin and Tsaï (5) investigated the definition of an expected modulus in function of fiber orientation. They defined two moduli El (longitudinal modulus) and Et (transversal modulus) as shown in figure 5. î ° Oļi s S¡t~*. ū ļ " Fias tic item Figure S: Description of the Halpin-Tsaï moduli. Injection moulding process leads with the mould filling and the fiber orientation as well. Mechanical 795 properties of fiberglass reinforced thermoplastics depend on many parameters such as : fiber/matrix adhesion, fiber rate, length/diameter ratio of fiber, fiber orientation inside the matrix. Data used in formula: Em = thermoplastic matrix Young Modulus Ef = fiberglass Young Modulus Vf= fiberglass volume rate L = fiberglass medium length D = fiberglass diameter Transversal modulus Et = Em* 1 + 2*NI* l-Nt*Vf Ef/ _j Nt = W7m - %»+2 Longitudinal modulus , 2*L*N1* Vf , 1 + EL = Em * l-Nl*Vf Ef/ _i Nl = Ef/ + 2* L/ /Em /D Random Haloin-Tsai model E = - *E1 + - *Et 8 8 Results show that the random Halpin-Tsaï modulus is veiy closed to experimental results (Standard tests NF T 57101). All the results are given in table 5 and studied in figure 6. It is very interesting because knowing the modulus value will enable us to have information about fibers orientation. Table 5: Experimental results compared to results expected from models. Tensile tests on saturated polyester PBT (3). Glass Test Random Halpin Halpin Fiber results Halpin- Tsaï EL Tsaï Et Volume (MPa) Tsaï modulus modulus fraction model (MPa) (MPa) (%) 0 11.12 6475 5244 8206 3467 17.65 8445 6928 11686 4073 33.33 14265 11468 20678 5942 || Young modulus progress in function of rate fiber! PBT tested under room temperature | 25 _ 20 - ^ ^ 15 y -» Longitudinal modulus E ■I % y* -a- Experimental results § 5 Random distribution of Hafein Tsaï ? * T ransversal modulus Et 0 0 10 20 30 40 Voumic rate fiber in % Figure 6: Evolution of the the fiberglass rate for diff EFFECT OF TEMPERATURE ON MODULUS - Specification sheet of the system impose high temperature rigidity tests. Glass transition of thermoplastic is an important parameter. Three points bending tests were made to measure the flexural moduli under different temperatures. Flexural modulus in function of the temperature is shown in figure 7. Flexural modulus in function of temperature] 10 "Jo 8 - S I I 6 - ' 20% gf Polyester ■a ' 20% gf PBT /AS A blend I 2 -SO 0 SO 100 Température ri Celsius d Figure 7: Evolut temperature. Results show that aromatic polyamide presents a glass transition around 80°C compared to saturated polyester and to the blend PBT/ASA which present a transition around 50 °C. Beyond glass transition, the modulus is decreasing very strongly and therefore the rigidity of the part. At this stage of the study it would be very interesting to resort to CAE in order to calculate the rigidity of different shapes. Many tools are available to optimise 796 rheological and mechanical behavior of technical part. Making calculations with this kind of software deals with a major problem: the need to know the accurate data of the materials. Working with fiberglass reinforced materials implies a delicate anisotropy to manage to in terms of modulus. This study will be the subject of a future investigation. CONCLUSION The addition of short glass fibers to a polymer greatly improves some of its properties. Thus reinforced polymer parts can be used in mechanical systems requiring stiffness and lightness. Yet mechanical properties depend broadly on the orientation produced by the injection process. It is interesting to find why and how fibers orient themselves. A new process, so called Gas Assisted Injection Moulding (GAM), would allow to obtain lighter parts with likely higher stiffness due to a better control of fiber orientation. Parts design is also expected to benefit from this new technology because of style freedom. Valeo is studying the cristallinity of the different interfaces (melt/mould and melt/gas) of parts injected with GAM process by using a Differential Scanning Calorimeter. At the same time it is interesting to focus on profit in terms of rigidity brought out by GAM. Results will appear in a future paper. REFERENCES (1) V.M Nadkarni and S. R. Ayodhya, The influence of knit-lines on the tensile properties of fiberglass reinforced thermoplastics, Polym.Eng.Sci, 33,358,1993. (2) J.L White, Editors: D.W Clegg and A.A collyer, Mechanical properties of Reinforced materials, Elsevier Applied Science Publishers. (3) G. Cervantes, "Etude de matériaux thermoplastiques renforcés de fibres de verre courtes destinés au moulage par injection d'une pièce d'essuie-glace automobile", rapport DEA, UPMC (Paris VI, France), 1994. (4) F.B Broussaud, V. Gossart, M. Thomas, Rapport technique de TONERÀ: Technique d'echos puisés appliquer à la détermination des modules, 1987. (5) M.L Benzeggagh, Micromechanics courses applied to composit structures, Université de Technologie de Compiègne, 1992. ACKNOWLEDGMENT The authors want to thank M.Thomas from l'ONERA who help them to carry out the ultrasonic measurements. 797