PETpla.net Insider 09 / 2013

BOTTLE MAKING 39 PET planet insider Vol. 14 No. 09/13 www.petpla.net What is FEM? An FE system will first divide the bottle shape into separate triangu- lar or square elements (FEs). The smaller these individual elements are, the finer the resulting mesh will be. The finer the mesh, the more exact the computed results, but computa- tion will also take much longer. With a view to increase the accuracy and/ or meaningfulness of the results while reducing the required computation time, the mesh is designed so as to be finer in certain areas of the bottle, e.g., in the shoulder, in the base zone, near the special design elements such as grooves, and in the lateral area. In addition to the data on the contour geometry of the bottle, each mesh element is assigned data on the wall thickness. Also, information on mechanical characteristics, such as modulus of elasticity, strength, and other material properties, as well as the laws governing the material, is entered into the FE system for the description of the behaviour of the bottle under mechanical and heat load. Entering and maintaining these data, rules, and tolerances in the FE system requires a qualified specialist in each company. What FEM can do FE systems permit the following statements: For the demonstration of the required top-load properties, an FE system will, in its most simple version, supply qualitative information based on a “red/green” (fail/pass) statement. In case of a failure, the position of any weak spots in the bottle can be identi- fied. The FE simulation of the top-load properties is also very well suited for a qualitative comparison of various shoulder geometries or the effect of grooves or beads in the lateral sur- face. FE simulations of the behaviour of a container under internal pres- sure will provide both qualitative and quantitative statements on the widening of the container, also near grooves or beads, as well as on the increase in the bottle height in excess pressure computations. In the case of hot-fill bottles, the deformation of the Please order your copy at the PETplanet insider book shop: https://www.petpla.net/books Bottles, Preforms and Closures A Design Guide for PET Packaging Second Edition by Ottmar Brandau 115,00 € 180 pages © Copyright Elsevier 2012 bottle caused by the vacuum that is produced when the hot contents cool down can also be simulated (see also Part 4). FEM also allows simulations of the behaviour of a bottle at higher temperatures. Apart from the top load and the internal pressure behaviour at elevated temperatures, the defor- mation of the base near the gate is of particular interest. These simula- tions will also reveal whether the gate will protrude from the bottom to such an extent that the bottle is not stable when standing upright. If one or several of these FE simu- lations show that the bottle does not yet meet the requirements, iterations involving changes in design elements, such as the number, position, and geometry of grooves or beads, can be used to simulate the required meas- ures until the bottle conforms to the specifications. Sometimes the FE simulation may show that a heavier preform is required. Within limits, a shifting of wall thickness values across the con- tour of the bottle is also possible. If the integrated FE system gives the go-ahead for the mechanical characteristics such as top load and internal pressure, this means that the wall thickness and the wall thickness distribution have been selected such that the bottle will withstand the antici- pated loads. Optimisation of these wall thicknesses with a view to gen- erating lightweight bottles will require the above-mentioned FE expert sys- tems or at least consultation with the FE specialist of the company. Such high-value FE expert sys- tems also permit a simulation of the shelf life of bottles for carbonated products. If bottle geometry, wall thick- ness, and – in the case of multilayer bottles – the thickness of the individ- ual layers are given, statements such as X% CO 2 loss after 90 days or 5% CO 2 loss after X days can be calcu- lated. Such simulations usually take two to three days, whereas measure- ments on real bottles would take at least four weeks. What FEM cannot do Despite the impressive capabilities of high-end FE systems, the temper- ature-dependent, nonlinear, elastic– viscous behaviour of thermoplastics such as PET has not yet permitted us to predict the exact wall thickness distribution in a bottle based on the input of preform, bottle geometry, and heating conditions. Nor is this input sufficient to determine the optimum adjustment of the heating panels in the oven system of the blow moulder. Yet, science and industry are already working on development projects to address these challenges. *This article was published in Bottles, Preforms and Closures, Ottmar Brandau, Chapter 1.5., Copyright Elsevier 2012