Shrinkage simulation of blow molded parts using viscoelastic material models

The utilization of simulation procedures is gaining increasing attention in the product development extrusion blow molded parts. However, some steps, like shrinkage and warpage, are still associated with uncertainties. reason for this on one hand a lack standardized interfaces transfer data between different tools, other complex time-, temperature- process-dependent material behavior used semi crystalline polymers. Using new vendor neutral interface standard transfer, analysis simple part investigated compared to experimental data. A linear viscoelastic model combination an orthotropic process- temperature-dependent thermal expansion coefficient prediction. good agreement observed. Finally, critical parameters models that strongly influence identified by sensitivity study. Der Einsatz von Simulationsverfahren gewinnt der Produktentwicklung extrusionsblasgeformter Kunststoffhohlkörper zunehmend Bedeutung. Einige Simulationsschritte, wie z. B. die Simulation Schwindung und Verzug, sind jedoch noch mit Unsicherheiten verbunden. Grund dafür ist zum einen das Fehlen standardisierter Schnittstellen für den Transfer Simulationsdaten zwischen verschiedenen Softwaretools anderen komplexe zeit-, temperatur- prozessabhängige Materialverhalten verwendeten teilkristallinen Polymere. Unter Anwendung eines neuen Schnittstellenstandards Datentransfer wird Schwindungsanalyse einfachen blasgeformten Bauteils untersucht experimentellen Daten verglichen. Für Schwindungssimulation ein viskoelastisches Materialmodell Kombination einem orthotropen prozess- temperaturabhängigen thermischen Ausdehnungskoeffizienten verwendet. Vergleich Messdaten zeigt eine gute Übereinstimmung. Abschließend erfolgt Sensitivitätsstudie zur Identifikation maßgeblichen Einflussgrößen auf Schwindungssimulation. molding process most common methods production hollow plastic parts bottles, cans, fuel tanks or large containers. During production, tube, which called parison, extruded. This parison then inflated against walls cooled mold. After certain cooling time solidifies can be demolded. Due relatively short cycle times manufacturability geometries, high economic efficiency ensured. because quality standards competition alternative processing methods, use computer aided engineering (CAE) increasingly attention. Various research groups presented approaches covering creation, inflation, finally warpage 1-7. In order achieve prediction, all steps involved must deliver sufficiently accurate results. These results need transferred from step following ones. Especially if software tools individual often not straightforward. required customer-specific very time-consuming cost-intensive. For inflation typical tests, prediction accuracy already quite good. Nevertheless, when it comes there One major problems influenced conditions. inside mold, possible only thickness direction prevented directions sustained blowing pressure (mold constrained) 1. Therefore, stresses build up. demolding, these lead shape deviations final (shrinkage warpage). Depending process-related wall distribution, local rates as well demolding temperatures vary considerably. general, higher result values vice versa. rate influences level stresses. reduction free due further ambient air. capture strong conditions, polymers such e. g. high-density polyethylene (HDPE), needs modeled broad temperature range. addition, has been observed parts, considered 8-10. Previous studies isotropic cover dependence behavior, but neglect addition description simulation, distribution during also have significant behavior. shows potential improvement. While practice lot paid design mold cooling, interaction usually neglected simulation. So far, no open literature consider flow conditions channels study, integrative approach presented. standard, so-called VMAP enhance operations sophisticated 11. computational fluid dynamics (CFD) improve thus extends workflow 7. latter, conjunction coefficient. stages study identify important influencing factors. consists four main Figure first involves manipulation (e. spreading clamping), inflation. Accuform/BSIM Version 2.5, finite element-based tool process, steps. self-developed interface, stretch ratios their orientations determined each element mesh 12. information necessary define Because meshes mapping routine required. MpCCI mapper Fraunhofer SCAI used. Once carried out, blown closed at air after analyzed using Simulia/Abaqus. flow-related heat coefficients (HTC), out Altair/AcuSolve. history subsequent mechanical determine part. As Abaqus analysis. deformed remaining residual occurred input top-load internal tests drop tests. Between tasks, amount exchanged, BSIM, AcuSolve used, exchange project, vendor-neutral storage was developed interoperability virtual workflows 13. could successfully integrated into fact always stored same structure offers more flexibility, even intermediate bulk container (IBC) operations, based [11]. Simulationsworkflow Vorhersage Verzug am Beispiel typischen Intermediate Bulk Containers unter Verwendung des Interface Standards Datentransfer, angelehnt Based previous simulative investigations, section 2 Also, will following, database, discussed. cylindrical made grade Lupolen 5021DX (LyondellBasell) measurements 7, 2a. geometry uniform thickness, minimized, allows independent investigation axial circumferential direction. To examine varying thicknesses, degrees stretching were investigated, Table Blow investigations [7]. Part measuring points (a), automated scan surface (b), distance measurement (c), comparison computer-aided (CAD) -model (d). Blasgeformtes Bauteil, welches Schwindungsuntersuchungen verwendet wurde Bauteilgeometrie Messpunkte automatisierter Scan Bauteiloberfläche Abstandsmessung anhand Scans dem Computer-Aided Design -Modell Test Point Degree [−] Thickness [mm] Cooling [s] Amount 1 1.5 30 4 60 3 90 5 2.0 6 7 8 measure direction, scanned GOM ATOS 3D scanner, 2b Several balls embedded provided marks comparing circumference cut middle test point, three period six days (processing shrinkage) months (post shrinkage), Furthermore, measured over 280 seconds camera (FLIR ThermaCAM SC1000). varies 1.4 3.3 %, expected, days, increase rather low. than explained stretched mostly leads molecule orientation With orthotropy increases. Results Ergebnisse Schwindungsmessungen sections, discussed detail. via meshed triangle membrane elements. diameter profile selected machine real parison. modeling, nonlinear K-BKZ finished, Together file mapped where T boundary, TU h several specified. determination challenging. complex. It depends pairing, roughness 15. Investigations bottles abrasive blasted aluminum pressures bar suggest 600 W m−2 K−1 1000 17. smoother surface, 1300 K−1–1800 inner whether accumulates continuously exchanged. case, convection whereas second forced convection. cools down case convection, value 10 assumed outer unknown, reverse engineering. water calculated globally turbulent 18. locally function velocity channels, Altair/AcuSolve channel tetrahedron At inlet channel, mass 0.195 kg s−1 experiments defined. outlet, zero-pressure boundary condition suggested. blocked outlets (via plug), defined zero (no-slip condition), 4. get field Navier-Stokes equation solved. Assuming flow, effort. reduce effort turbulence model, Spalart-Allmaras physical properties taken 19, 20. coefficients, method non-dimensional (HTC-method 3) 21. calculation dimensionless parameters, so additional necessary. determination, saved MpCCI-Mapper. Boundary Randbedingungen Strömungssimulation. Like Abaqus. (cooling under constraint), freedom nodes fixed. change temperature, changed, shrink freely days. investigate gravity load applied demolding. quasi static (Abaqus visco step) density both analysis, directly imported. densities mapper. characterize (Lupolen5021DX, LyondellBasell), dynamic performed samples, samples stamped square bottle evaluation Proteus 6.1 Netzsch, master curves created range – 20 °C 120 reference 23 WLF time-temperature-superposition. specific elastic modulus low, corresponds studies, 8, 9. dynamisch-mechanischen Analyse Proben, Die Probenentnahme erfolgte aus Vierkantflaschen Axial- Umfangsrichtung. Since evaluate up about °C, become too soft, curve extrapolated logarithmically low stiffness 19 Prony total curve, 6a. account equation. creation. Material Comparison series (AD) (CD) ratio (SR) (b). (a) obtained Materialdaten Schwindungsanalyse. Prony-Reihe orthotroper Wärmeausdehnungskoeffizient axialer Richtung Umfangsrichtung Verstreckverhältnis 1,5 2,0 experimentelle Masterkurve [7] entnommen. link Starting 15, 16, modified results, 3.1. Isotropic above 1, means either equal. below increased constant factor maximum (circumferential CD) reduced minimum (axial AD). shift fitted correspond experimentally values, 6b. better overview out. variations sections. described 3.2, just considered. practice, cycles needed starting changing ensure stable process. investigated. without steady-state temperature. simplification, dependent according 16 experiments, experience user. many cases, 80 %–90 % achieved. estimate error shrinkage, +/− one-element-model 3.2.4 refer time, demolded, creep 23. type (LyondellBasell). parametric 3.2.3 understanding considering parameters. (different thicknesses) outcome mm points, measurement, half. scanner half thickness. visualize percentage deviation Even larger partial areas, area (square) Percentage components (TP) 2, 4, Prozentuale Abweichung Wanddickensimulation vermessenen Bauteilen Versuchspunkte Tabelle illustrate 8. backwater areas (dark areas) zero. average m 24000 K−1. 40000 big difference global coefficient, 9370 4.2. Velocity (HTC) (b) (c). Geschwindigkeitsfeld lokale Wärmeübergangskoeffizienten Netz Abkühlsimulation gemappte while difficult, (section 3.2.2), measurements. biggest point image Ergebnis Thermokamera wurden area, took place, 1–4 5–8, 10. oriented Illustration 3, Darstellung ermittelten Verstreckgradverhältnisse able match little bit underpredicted qualitative 5, negligible. contrast automotive 3. long thin seem subject warpage. therefore Gegenüberstellung Simulationsergebnisse Schwindungswerten. 4.1, seven heated up, small. concluded five cycles, equilibrium state reached, seventh thermocouples, placed close cavity. simulated underestimated. parameter computation (about ten cores AMD Ryzen9 3900X). Considering small changes temperatures, sufficient model. Temperaturverteilung Werkzeugform Versuchspunkt nach Kühlzyklus Kühlzyklen two positions (T1 T2). sensors upper component system had reached steady ersten zwei Temperaturfühlerpositionen Temperaturfühler im oberen Bereich Mitte platziert. Messung erst durchgeführt, nachdem System stabilen Zustand erreicht hatte. siebten Zyklus Simulation. resulting evaluated gradient taking 15 across less lower 14a. Consequently, 14b. cavity high. itself high, probably effect. simplified special component, consideration general seems play subordinate role. cavity, cannot ruled temperatures. Variation Influence Abkühlsimulation. Einfluss gemittelte Entformungstemperatur Bauteilschwindung importance, considerably pressure, pairing. According 17, reasonable 1800 initial 37 15a. effect 17 (test 1), 15b. its great influence, should examined closely pressures. Artikel Blasformwerkzeug. variation 16. %–28 0 %–24 depending point. whose through extensive errors hardly avoided. Thus, aware designing Wanddicke. thicknesses times. allow insight respect illustrated functions 17a. marked circles. clearly visible approx. relationships inverted. show they °C. occur faster cooling. hand, much slower relax. Simulated Simulierte Schwindungswerte Abhängigkeit gemittelten Entformungstemperatur. verschiedener Wandstärken Wandstärke zu Zeitpunkten Entformung periods, 17b. 50 hour past would shrinkage. observed, diminishes time. Unfortunately, database lacking conclusive effects. work, within workflow. Local dissipation channels. polymer polyethylene. predicted, deviate qualitatively quantitatively. influence. huge thicknesses. chosen highly Additional comprehensively possible. immediately provide valuable information. differences work funded German Federal Ministry Education Research (BMBF) ITEA3 cluster European initiative EUREKA (Funding Sign BMBF 01|S17025A–K). We thank Graduate Institute TREE Hochschule Bonn-Rhein-Sieg University Applied Sciences financial support. Open access funding enabled organized Projekt DEAL.