9 Computer-Aided Analysis

9 Computer-Aided Analysis Computer-aided analysis (CAA) of a variety of plastic processes is available For the purposes of this book, CAA includes a finite-element analysis of what may be happening in an injection mold during the molding cycle In spite of the use of a computer for the analysis, this is not an exact science Many assumptions are involved in the computer algorithms The program operator must make yet more assumptions Thus, the end result of the analysis can follow the well-known computer admonition: garbage in, garbage out Nevertheless, a skilled computer operator can arrive at fill rates, filling and cooling times, shrinkage and warpage values, fill pressures, and pressure distributions that are more accurate than those that can be estimated by the most experienced mold designer or builder There are only a few programs on the market that qualify as good analyzing programs.[64] Among the longer term players are Plastics & Computer’s TMconcept® family of software tools and Moldflow’s Flow Analysis family of programs There are other companies that offer analysis packages If the intent of the end user is to check a box that says the analysis was performed without using the analysis to optimize the process, any program will Any good analysis software should yield results that are in line with what you expect when you model a very simple part without using “fudge factors.” If you have to use fudge factors to make the analysis work out as expected, how can you trust the analysis when the part is complicated? Giorgio Bertacchi of Plastics & Computer, Inc., says, “We contend that no computer program can compensate for a user’s inexperience In the hands of nonprofessionals, even the best models, based on process fundamentals and using transparent, automatic modeling, carry the inherent risk of producing erroneous results and causing costly mistakes.” For any analysis, someone with a lot of experience should review the results If the results appear to be out of line, then a careful review of all assumptions and inputs to the program are appropriate Before accepting the results, a logical reason for the unexpected results should be found 9.1 Capabilities of CAA Injection molding is an art of compromise What are the objectives that you are trying to achieve? If a © Plastics Design Library fast cycle is the objective, then better cooling may be the purpose of the analysis If holding tighter tolerances is the objective, then longer cycles or a different resin may be indicated If the molding project has a small window of moldability, some changes might be advisable to avoid excessive mold maintenance such as repairing gate wear or cleaning minerals out of the water lines For example, how you clean out the water lines of a mold that is built with “conforming” water lines? Conforming water lines are water channels that are formed into a mold insert by one of several processes whereby the water lines follow the molded part profile at a constant distance from the mold surface These water lines are not straight and are not drilled They may have any number of twists, turns, or other convolutions that defy mechanical cleaning The premiere analysis systems that use finiteelement methods consist of a number of modules Each module simulates a different portion or aspect of the process For example, one module will take a CAD (computer-aided design) model and mesh it for analysis Coupled with that module are modules that analyze the filling and the packing/holding phases of the process Other modules predict the resulting shrinkage/warpage or final shape of the part, or remove some simplifying assumptions about the cooling capabilities of the mold In addition, there may be modules to analyze special subsets of injection molding like gas-assisted molding or injection compression molding Decision support modules may also be available that offer quick approximations to help guide the detailed analysis process and identify the hurdles and challenges presented by a particular application Some of these modules can be used even before a detailed CAD drawing is completed and can be used to help guide design decisions to ensure a robust process and part quality These modules offer estimations regarding the difficulty of filling the part, attainable tolerances, shrink rate, machine capability determination, etc In addition, these programs typically look at the economic impact of various design decisions and present a detailed engineering cost estimation The costing portion should help with decisions on mold features such as recommending the number of cavities and runner type, as well as molding machine capability requirements, and production planning issues Ch 9: Computer-Aided Analysis 128 Each of these CAA programs requires good knowledge of the molding process and of the assumptions made in the computer analysis programs in order to obtain reasonably accurate results Probably the most basic assumptions deal with the relationship between pressure, temperature, and volume These relationships are well known and documented for relatively slow cooling rates, say five degrees per minute The relationships between these variables at cooling rates of perhaps hundreds of degrees per minute are not commonly available Therefore, certain assumptions are made about these relationships when analyzing mold filling, cooling, and shrinkage These three variables are the most prominent of the variables to be considered, but there are approximately thirty total variables Most finite-element–based analysis programs use what are called midplane analysis techniques Midplane analysis involves making a model of the midplane of part That midplane surface model is then meshed with either triangular or quad plate/shell elements The appropriate thickness property is then assigned to each element Once the mesh is generated and the thicknesses defined, the gates and runners are typically added and defined The gates and runners are normally one-dimensional elements with length and diameter or size properties In some programs, gate and runner elements may have special element types to better define their flow and heat-transfer properties (for example, hot runner, cold runner, or insulated runner) Calculation times will vary by program and will depend on the part-flow configuration Most analysis output consists of pictures and graphic data that indicate the flow-front at any time during the filling process, and the temperature, shear stress, shear rate, frozen skin, and pressure distribution at any instant during the process Fully dynamic programs, like Plastics & Computer’s TMconcept® programs, recompute all the field variables in each element back to the origin of flow at each interval of time; other programs assume that once an element is filled, the conditions in that element only change on a time-dependent basis (that is, the shear stress stays the same, but the temperature drops due to time-dependent heat transfer) Due to the latter assumption and the assumption of “fountain Figure 9.1 The injection pressure and flow-line distribution that result from the use of sequential gating.[61] (Courtesy of Plastics & Computer.) Figure 9.2 An analysis of thick-walled parts where highpressure gas is used to fill out the mold The gas creates voids in the heaviest sections so that the parts are hollow This minimizes the amount of plastic required, creates hollow parts, and minimizes sink marks.[61] (Courtesy of Plastics & Computer.) An analysis may result in the use of a smaller molding machine for large parts by optimizing the gate location to lower injection pressures An analysis can help size runners and gates in family molds to ensure that all cavities fill at the same time It can help arrange gates and flow patterns to minimize the tendency for cores to shift under injection pressure It can help a mold designer position and time sequential gates (see Fig 9.1), so that as the flow-front passes a new gate, it opens, thus avoiding weld lines and minimizing flow distance and cavity-pressure differential Gas-assist injection molding simulations (see Fig 9.2) help determine the correct size of the gas channel, the shot size to be used, and the process conditions to ensure the desired size of the voids left when the gas displaces the plastic in the heavier sections such as rib intersections 9.2 Limitations of CAA Ch 9: Computer-Aided Analysis © Plastics Design Library 129 flow,” some programs can erroneously identify the areas far from the gate to be hotter than the areas near the gate The metric system is preferred in CAA for molding plastic Round-off errors can result in division-byzero errors more often with inch units than with millimeters (A millimeter is about 1/25th the size of an inch.) This is primarily a problem in small parts Shrinkage can vary widely It is influenced by many factors already discussed, but the shape of the part and its constraints while in the mold are significant Some of the simplified decision support programs, like Plastics & Computer’s MCO (Moldability and Cost Optimization) programs, not generally consider such restraints to shrinkage They assume that the parts are allowed to shrink to the extent that molding conditions predispose them In other words, molded parts that are constrained may appear to shrink much less (or more) than the analysis indicates due to warpage caused by differential shrinkage and physical constraints Unlike finite-element based shrinkage/warpage programs, MCO can consider shot-to-shot and cavity-to-cavity variabilities to come up with an anticipated range of shrinkage so that attainable tolerances can be more effectively considered Finite-element shrinkage/warpage is a simulation and cannot consider the shot-to-shot and cavity-to-cavity variations However, it does consider warpage and the user can apply constraints To consider the impact of variations in the process, multiple analyses under different conditions need to be run This process can be very time-consuming and will not account for the cavity-dimension tolerances due to toolmaking Most analysis programs today assume that there is adequate venting, so no backpressure is considered during the filling stage As all molders know, inadequate venting can significantly affect the moldability of a part, and the filling pattern None of the current analysis programs have specific result displays to addresses surface finish imperfections Some programs provide displays indicating weld-line location, but these should be used with caution and verified by looking at the flow-front development since there are frequent reports of incorrect indication, and the analyses not offer any indication regarding the potential severity of the resulting surface or structural problems It is generally recommended that weld-line formation and integrity can be evaluated by interpreting the flow pattern and melt conditions at the time that the flow fronts meet Other phenomena like surface roughness from inadequate venting, moisture, or stick-slip skin folding are not ana- © Plastics Design Library lyzed, although users with extensive molding experience may be able to anticipate some of these by interpreting changes in the field variables (temperature, stress, pressure, etc.) during the molding process Some programs claim to predict the depth of visible sink marks (see Fig 9.3) 9.3 Selecting a CAA Program There is a tendency for people to accept the output of a computer program as an error-free fact, forgetting that an imperfect human wrote the program and the operating system The computer analysis of plastic flow, cooling, and shrinkage within a mold requires consideration of many variables, some of which change from moment to moment during the molding process and cannot be predicted in advance Other parameters vary with the age and condition of the mold and molding machine Therefore all analysis programs must make assumptions What these are and how they are addressed in the computer program affect the end results The CAA results should not be based on faith but rather be subjected to intense scrutiny Before selecting a program or accepting the results of an analysis, there are certain questions that will help determine its accuracy and validity First of all, the user should be aware of the assumptions that are built into the analysis program Carefully determine what these assumptions are and how they will affect the analysis results Figure 9.3 Filling pressure distribution and potential sink marks.[61] (Courtesy of Plastics & Computer.) Ch 9: Computer-Aided Analysis 130 Consider how the program handles branching flow into the mold Even a single-cavity mold has flow branching as the flow moves away from the sprue or gate through one finite element and spreads out into two or more other elements Does the program assume a constant flow rate? Does the flow rate change in each element as the flow diverges from the gate? Does the program consider a modern molding machine’s ability to vary the flow rate as the molding cycle progresses? Do the analysis results show that flow advances faster in thick sections when compared to thinner sections? To put it another way, does the flow-front advance inversely when compared to the resistance to flow? Consider a simple mold containing two cavities of vastly different volumes but with a common runner, gate, and cavity thickness Does the program predict that they will fill at a different time, as it should? (See Fig 9.4.) How about a mold with two cavities, each with the same flow length but with different cavity thicknesses? Does the program predict that the cavities will fill at a different rate and pressure? How does the program handle shear rates? Shear rates will vary depending on skin thickness as the mold fills Some programs have assumed that no solid skin develops as the mold fills so that the maximum shear rate occurs at the mold surface The analysis program should predict the different skin thicknesses and temperatures that result from very long, slow injection cycles, and short, rapid injection cycles How can you verify temperatures calculated and how does the program deal with crystalline materials? One simple test is to determine actual no-flow condi- tions within a test mold by increasing packing or holding time until the part-weight stops increasing, while carefully documenting all parameters Determine one set of conditions for an amorphous material and another set of conditions for crystalline materials Compare the results with the analysis program If the analysis program fails to accurately predict the no-flow temperature, its other results are suspect Are cross-section temperature predictions reasonable? (See Fig 9.5.) It has been established that temperature profiles through the thickness of a part vary widely depending on flow rates At high flow rates, a shear-heating temperature peak occurs near each wall of the cavity At low flow rates, the temperature peak near the wall fails to develop because there is little shear heating Testing the analysis program’s temperatureprofile predictions at high and low flow-rates should show a peak near the wall at high flow-rates and no apparent peak at low flow-rates Does the program consider and recalculate conditions in each element based on the influence of other elements as time progresses? As resistance to flow increases in one area, is the flow shifted to other areas that are experiencing lower resistance to flow? Does the program predict plastic temperature rise based on increasing shear rates? Any flow analysis program should give results that are consistent with an experienced molder’s observation of the real world If the predicted results are inconsistent with expected trends, then the analysis should be suspect Figure 9.4 The effects of adjusting runner size to ensure that both cavities of a two-cavity mold complete the filling sequence at the same time [61] (Courtesy of Plastics & Computer.) Figure 9.5 Temperature distribution and temperature crosssections in a mold.[61] (Courtesy of Plastics & Computer.) Ch 9: Computer-Aided Analysis © Plastics Design Library 131 9.4 Customer Requirements The hardest job for the person making the analysis may be to get the person requesting the analysis to precisely define his goals If a “complete analysis” of a part would cost $10,000, the actual requirement might be much less if the exact purpose of the analysis is defined For bids on analysis, include a rendering or drawing of the part and a careful description of exactly what analysis is desired and what your goals are What are the purposes of the filling analysis? Is it to size runners and gates? Perhaps it is to determine if the part will fill? Is the shrinkage of the part of primary concern? Is distortion due to warpage a primary concern? How can the cooling and cycle time be improved? Can the quality of the part be improved? (See Figs 9.6–9.8.) What can be done to minimize size variations? What can be done to minimize or eliminate sink marks? By moving the gates, can the part be filled on a smaller machine? Is the available machine adequate from the standpoint of shot size and clamp capacity? Can you hold the tolerances requested? Do you need to consider a different resin? Do you need to consider all available manufacturers and grades or can you limit yourself to a single manufacturer’s specific resin and grade? What are the operating conditions of the finished molded part? Is it going to be used in Alaska or Saudi Arabia? Widely differing end-use temperatures can cause parts to be out of tolerance due to the coefficient of thermal expansion differences in mating parts of dissimilar materials How are the parts inspected, and at what temperature? The customer should Figure 9.6 A separate gate at the root of each fan blade, fiber orientation, and distortion in a shrouded fan.[61] (Courtesy of Plastics & Computer.) © Plastics Design Library carefully consider these questions and others, and define carefully what he expects of the analysis Even though a resin may meet a set of specifications, variations in flow and shrinkage between different manufacturers can throw a part out of tolerance What are the manufacturing issues? One example is that of a medical tray of Ultem which was analyzed The original question was “Can the tray be molded with two gates?” The analysis showed the tray could be molded, but at a pressure near 20,000 psi Most machines are capable of this pressure, but what of the clamp force required to keep the mold closed? The injection pressure times the projected area of the part indicated the need for a clamp pressure of more than twice that available to the molder Redesign of the gating allowed the part to fill with three gates and within the clamping capacity of the molder’s machine Figure 9.7 An analysis of a molded tray showing improved distortion and pressure distribution using two gates instead of one.[61] (Courtesy of Plastics & Computer.) Figure 9.8 Distortion improvement that results from using two center gates instead of two edge gates.[61] (Courtesy of Plastics & Computer.) Ch 9: Computer-Aided Analysis 132 Who normally requests an analysis? It could be anyone involved in the design and production process The designer, the engineer, the molder, the moldmaker, and the end user each has an interest in producing a satisfactory part The best arrangement is for all of these people to be on the same team, working together and using the analysis software to optimize the design of the part, the design of the mold, and the molding conditions to maximize production and profit That way, the expertise of all the team members is utilized to find the best set of compromises available When used correctly, the analysis serves as a virtual mold trial, where trying different options is relatively cheap, easy, and fast It helps improve communication between the team members and, therefore, can make design review meetings more productive and allow the team to push the limits of the standard practices Anne Bernhardt, of Plastics & Computer, Inc., (who sell the TMconcept® line of software), suggests that the least experienced designer or engineer with CAD knowledge run the programs and “punch the keys,” with the more experienced team members determining the issues, desired results, alternatives to try, and helping in the interpretation of the results This helps less-experienced members of the team rapidly learn the molding process and problems that occur in real-world production while still being a valuable member of the team Through the way their menus are written and some of the results are presented, most programs have some tools to help guide the options that are considered The part designer is the member of the team that can usually answer questions about part modifications He learns from the analysis which features cause problems, and that improves his future designs The moldmaker and molder better understand the designer’s intent and requirements, and also gain valuable insight about each other’s strengths and constraints Management gains a valuable tool to understand how to maximize production and profit 9.5 Management Tools Simplified programs that offer very fast calculations, simplified inputs, and consider economics are important tools for decision support and project management These programs should let you evaluate the viability of a project at the initial concept stage and refine the inputs and analysis as decisions are made Ideally, you should also be able to use these tools to evaluate improvement options of existing production Ch 9: Computer-Aided Analysis Unlike standard simulation programs, these tools calculate costs, not require detailed CAD drawings, and some consider process variability and machine capabilities These programs use a lot of simplifying assumptions As a result, many believe them to be inferior to detailed simulation programs; however, in many cases they offer more “bang for the buck.” Because decision support programs are very fast, and require very few inputs, they make it possible for the product development team to consider many more options than without them The economic impact of changing resin and manufacturing constraints can be considered, as well as the economic incentive to overcome limitations (mold size and thickness problems, excess tonnage or shot size or residence time, clamp stroke for deep-draw parts, recovery time, etc.) or change part or quality requirements Decision support programs are not meant to replace simulation programs, rather they help guide the design process by helping the team select the best options and focus engineering resources on the aspects that are likely to cause problems in production Some programs are limited to estimating the ability to fill the part, the associated clamp requirements, and an estimate of mold-closed cycle time Others also let you evaluate economics and costs, the total cycle, including machine actuation time, tool size and cost, general cooling requirements, attainable tolerances, and other factors An important additional benefit to decision support programs is that they provide the basis for establishing a methodology that ensures that all aspects of the application are considered early in the project The early identification of features that are difficult or costly to achieve enables the team to focus on design alternatives in these areas while changes are relatively inexpensive to make Decision support programs like Plastics & Computer’s TMconcept® MCO (Moldability & Cost Optimization) programs estimate cycle time, processing conditions, and required gate size based on the resin, a simplified description of the part geometry, and tolerance requirements Economic factors such as optimum numbers of cavities, machine selection, and batch size can be optimized based on machine availability and capability, production requirements, part quality requirements, and costs The program also determines the resultant yields, production-planning data, and the finished part cost A plant database with hourly rates and machine capabilities reduces data entry The program also helps identify factors that could limit pro- © Plastics Design Library 133 ductivity and/or increase costs MCO also has the capability to add markups, as well as the cost of inserts, secondary operations, and transportation costs, to come up with a sale price for the finished part 9.6 Filling Analysis A filling analysis simulates the filling phase of the injection-molding process In other words, it covers the time from the initial introduction of melt into the mold until the instant that the entire mold is filled with resin Filling analysis requires a definition of the part or mold geometry, a resin database, and molding conditions Based on the way the geometry is defined, there are four major categories of filling analysis on the market today See Fig 9.9 Lay Flat or User Defined The oldest form of flow analysis, this method is sometimes called a 2D (twodimensional) method The part is defined in segments that approximate how you expect the part to fill Various segment geometries (radial, rectangular, round, etc.) are available to describe the filling pattern in the part and runner system This method requires a lot of user knowledge and understanding of what the most likely filling pattern will be In recent years, this method has been most commonly used for runner sizing and balancing This method is particularly good in small, single-gated parts Mold Masters offers a program of this type called FillPlus™ This program starts with an expert system to help the user select the correct com- Figure 9.9 Several filling-analysis program results Notice the flow hesitation in the upper left corner where a “living hinge” is creating an impediment to flow.[61] (Courtesy of Plastics & Computer.) © Plastics Design Library ponents from their product line, and then completes the flow analysis for verification It can also check for the number of shots required for a color change Midplane FEA The most common flow analysis is the midplane FEA (finite-element analysis) method, which is sometimes called a 2½ D method The part is described as 3- (triangles) or 4- (quads) noded elements on a midplane of the part These elements are then assigned thickness properties to define the part volume Examples of this type are Plastics & Computer’s faBest® programs and Moldflow’s MPI (Moldflow Plastic Insight) programs This type of program is the most thoroughly tested and widely used Although excellent for most injection-molded parts, it is difficult to use in modeling parts with very thick wall sections where it is hard to determine a midplane, and in very small parts, or parts with a lot of detail Determining the midplane can be time consuming Many CAD programs and some plastics analysis programs have midplane generators; however, many users report that they work very poorly For most medium- to large-size parts, using either outside surface generally works fine if there are no significant features on the other side One of the most important aspects of the meshing is to ensure that there is “connectivity” between the elements Without connectivity, the material can not flow between the elements Most mesh generators have utilities to check and repair connectivity Solid FEA Also called 3D, this is the newest type of analysis An example is shown in Fig 9.10 These are true 3D solid element programs where the solid CAD model is broken into bricks, hexahedrons, or tetrahedrons These programs are excellent for very thick-walled parts, small parts, and fiber reinforced parts One of the major drawbacks of these programs is the excessively long calculation times required by some Current commercial programs in this category include Plastics & Computer’s faSolid™, and Moldflow’s MPI/3D Figure 9.10 A representation of a solid FEA analysis during the filling operation.[61] (Courtesy of Plastics & Computer.) Ch 9: Computer-Aided Analysis 134 Dual Domain FEA This method is patented by Moldflow and used exclusively by them Their MPI/ Fusion product line uses this method It is a clever way to automate the process of meshing a solid model in STL format, but it creates new problems Initially, this meshing technique resulted in physically incorrect flow patterns in the presence of simple ribs on flat surfaces Some solutions have been added to help resolve these problems, but they increase the meshing and calculation times, and the quality of results seems to be more sensitive to the mesh than those of the midplane meshes The resin database for all filling analysis programs includes thermal and rheological properties Some programs, like Plastics & Computer’s faBest® and faSolid® also require the latent heat of crystallization for crystalline and semicrystalline materials Many software suppliers include menu-driven programs that allow the user to enter her or his own materials into the database since it is impossible and impractical to include every grade available on the market Processing conditions are generally entered through menus when an analysis is set up These inputs include selecting the melt entry location, the resin, the fill time or injection rate, the injection profile, melt temperature, mold temperature, and the V/P changeover point (switch from volumetric control to pressure controls) In most cases, the analyses will use the assumption of a uniform, assigned mold-surface temperature Some programs allow specific mold temperatures to be assigned to the “a” and “b” side of certain elements, or for the mold temperature to be refined by integration with the cooling analysis, discussed below, Sec 9.9, and in Ch The results of a filling analysis include the pressure required to fill the cavity, opening forces generated by the injection pressure on the projected area of the mold, and animated views of the progress of filling the part, as well as the distribution of field variables during the process Field variables typically include temperature, pressure, shear stress, shear rate, frozen skin, and flow orientation Plots of the flow rate and injection pressure at the melt entry-point and of the progression of the field variables can also be displayed In addition, each supplier offers a variety of displays aimed at helping the user evaluate the results or identify things like the location of weld lines Evaluation of the advancing flow-front shows the filling pattern and makes it possible to predict weldline location, the last point to fill, and other locations of potential air entrapment where vents will be needed The quality of weld lines can be evaluated by looking Ch 9: Computer-Aided Analysis at the melt temperature, shear rate, and frozen skin as the weld line is formed The following are general guidelines for evaluating the various filling-analysis result displays Cooling Time This is the time required for the center of the element to reach the freezing temperature of the resin (as specified in the database) starting at the end of the filling of the part This time is used as a reference to set the cooling time It normally represents the maximum cooling time since some parts can be ejected with a partially hot core Frozen Skin The frozen skin is the percentage of material frozen during the filling of the part For example, 10% frozen skin on a 3-mm thick part means that the frozen layer in each side is 0.15 mm This variable is essential to optimize the molding conditions and is a very interesting index to use for judging the quality of the part because it measures the frozen orientation The allowable amount depends on the type of material The frozen skin is very important for parts with very thin wall thicknesses molded with crystalline materials This variable may also be important for large parts (such as automobile bumpers) needing very long filling times, and where the heat transfer to the mold can be higher than the heat dissipation Isochrone This view shows the evolution of the filling phase since it is a multicolored picture of the advancing flow front Each color corresponds to a different short shot with its time No-Flow Time No-flow time is the time it takes for all layers in an element to reach the no-flow temperature of the resin (as specified in the material database) starting at the end of the filling of the part It gives the first indication of the packing of the part (the theoretical maximum holding time for each element) Opening Force The opening force is the force acting on the mold that needs to be opposed by the molding machine clamping force It is generated by the filling pressure acting on the projected area of the model It can be determined at various instants during the injection time In cases where the pressure for the subsequent holding phase is higher than the pressure required for filling, the final view must be carefully evaluated In fact, during the pre-holding phase after the V/P change, particularly if the melt compressibility calculation has been activated, the final pressure distribution might not be equalized in the whole part and give an underestimation of the clamping force required in the holding phase It is recommended that a holding/ packing analysis be done in all cases where the clamp- © Plastics Design Library 135 ing force during the holding phase is a critical requirement Orientation Orientation is an indication of the main flow-stream in each element As with the other variables (e.g., temperature, stress), it is calculated at each time-step during the filling phase Orientation is used for a better understanding of the filling pattern in order to judge potential causes of warpage The examination of the orientation’s velocity vectors becomes very important for materials with anisotropic shrinkage, like all the glass-reinforced resins Pressure Distribution The pressure distribution indicates areas of overpacking, which can cause differential shrinkage and consequent warpage Fillinganalysis programs perform the calculation of the initial holding phase for all flow paths that are filled prior to the end of filling the entire mold Note that at 100% of filling, it is common to find differently packed areas that are assumed to be identical It happens because of minor differences in the mathematics of the calculation due to geometry (for example, the position of symmetrical nodes not being exactly symmetrical), and the convergence of field variables (local temperature, pressure, etc.) These “errors,” which not play any significant role in the evolution of flow but cause minor distortions in the flow front, seem much more evident in the pressure distribution just near the completion of the filling phase Since this situation lasts just for an infinitesimal time, it cannot be considered as overpacking When in doubt, look at the view saved just before the completion of flow (for example, the V/P change point) Actually, this phenomenon of unbalancing near the 100% filling occurs also in practice, and it is the reason why a safety factor in clamping force is usually required to avoid flashing In injection molding, it is always possible that a minor difference (in this case, of local temperature or cavity thickness) can cause apparently identical areas to reach pressurization at slightly different times Shear Rate This is the gradient of the difference in velocity between adjacent laminar layers within the flow channel, divided by the distance between them The maximum shear rate across the thickness of the segment is shown See the shear-stress considerations Shear Stress This is the ratio between the shear force, which drives the flow, and the area resistant to flow It is a function of the material viscosity and the flow rate The stress displayed is the maximum shearstress across the thickness of the element at various instants during filling During cooling, part of the stress at the end of the filling relaxes, but a residual stress © Plastics Design Library remains frozen-in and will be one of the causes tending to distort the part The shear stress should not go above a specific limit that is a function of the type of plastic Typically, in the part, it should not exceed 0.3 to 0.7 MPa This value is also a function of the temperature and frozen skin In fact, high stresses can be found either in situations of high velocity and hot material, or low velocity and cool material The latter occurs due to high material viscosity Because the level of stress, which relates to part quality, is basically the stress that can be frozen in the part, it is evident that one can accept much higher values of stress in the first case, since it will have more time to relax thanks to the higher material temperature, than in the second case Temperature Temperature displays represent the average temperature of the material across the thickness of each element Temperature can be obtained at different time intervals and at the end of filling To obtain high-quality moldings, the temperature difference in all elements describing the part should be in a narrow range It requires that the heat lost by conduction to the cold mold-surface be compensated for by the heat generated by friction The maximum allowable difference depends on the plastic See Fig 9.11 A temperature rule-of-thumb: at the end of flow, the material should not cool down more than 15° to 20°C when compared with its typical average value Whenever possible, it is desirable to heat the material about 10° to 15°C by friction in the runners In very difficult filling situations, one can even accept heating the material by 10° to 15°C due to friction in the part near the gate Figure 9.11 Several possible outputs of an analysis program, including temperature, in a molded part at a particular time.[61] (Courtesy of Plastics & Computer.) Ch 9: Computer-Aided Analysis 136 9.7 Packing and Holding Simulation Holding and packing analysis programs extend the filling analysis calculations through to part ejection The inputs include the holding pressure (which may be profiled), holding time, and the cooling time The output of these programs include the distribution of pressure, frozen skin, shear stress, temperature, density, and volumetric shrinkage in the part during this phase of the process Some programs also include estimations of the risk of sink marks (Fig 9.3) throughout the part One of the most important graph outputs in this kind of analysis program is the plot of the entering mass over time This helps ensure that gate freeze-off is achieved prior to release of the holding pressure This is also one of the few variables that is relatively easy to verify The hold (or pack) time is the duration of time that melt pressure is maintained on the melt within the mold cavity This portion of the cycle typically accounts for less than 5% of the part weight but is critical in determining the final part density, part weight, and therefore the shrink rate This is especially critical in semicrystalline resins that go through a phase change that results in a relatively significant change in density The pressure can only be maintained as long as the gates and runners remain unfrozen If the holding time is too short, and the gate is still unfrozen, melt may flow back out of the cavity, causing high shrink rates and more shrinkage variability Similarly, if the runners have high levels of frozen skin, the pressure loss in the runners may limit the ability to pack the part Holding/packing modules are typically considerably less expensive than the filling analysis modules They are strictly an add-on module and fundamentally consist mostly of extending the filling calculations 9.8 (or more) points on the molded part and offer a variety of displays to mimic a wide variety of dimensional evaluation methods such as flatness or deviation from a defined plane, out-of-round conditions, etc Some include special views to help find a nominal shrinkage rate for tool making Shrinkage/warpage modules are generally quite expensive and calculation times generally take longer than the calculation times of filling or packing analysis 9.9 Cooling Analysis Cooling analysis modules allow an accurate determination of the effectiveness of the mold-cooling system at maintaining the desired mold temperature, avoiding hot spots, and meeting desired cycle time These programs are generally integrated with the filling and packing/holding modules They perform transient dynamic heat transfer analysis aimed at either determining the required cooling time for selected elements to reach a specified center-line temperature, and/or they predict the temperature distribution at the end of an assigned cooling time See Fig 9.12 These program modules should include the model of the cavity or a means whereby the cavity and mold may be modeled, and methods for modeling cooling lines, fountains, baffles, or any other cooling configuration The program modules should also have options to include a number of inserts with different heat-transfer properties In addition, identification of circuit loops should ideally be part of the calculation setup, which will also include the water temperature and flow rate Shrinkage/Warpage Simulation Differential shrinkage, residual stresses, and residual thermal stresses contribute to warpage The amount of distortion is also affected by the overall rigidity or inherent mechanical constraints due to part geometry Shrinkage/warpage modules are extensions to the filling/packing/holding analysis that predict the final shape of the part They are in fact a strain analysis, where the stresses have been determined during the previous analyses Shrinkage/warpage modules predict the direction and magnitude of warpage The program should be able to predict the linear shrinkage between any two Ch 9: Computer-Aided Analysis Figure 9.12 Cooling analysis, with cooling cross sections in the upper right corner.[61] (Courtesy of Plastics & Computer.) © Plastics Design Library 137 Results should include a variety of views to help evaluate the quality and uniformity of cooling: temperature distribution through the mold, cavity surface temperature, and the temperature difference between the “a” and “b” surface, water temperature distribution, etc 9.10 Costs The analysis tools discussed in this chapter are available in modules, each of which adds to the cost and capability of the analysis system At the present time, all injection molding software suppliers offer perpetual licenses Some also offer annual licenses or software leases Perpetual license costs vary widely between modules and suppliers The simplest individual modules can cost as little as a few hundred dollars, while the price for complete suites (filling, packing, shrinkage/warpage, and cooling) can vary from $35,000 to $150,000, depending on the supplier and quality of the software If you expect to receive technical support and to keep current with new developments in the software, an annual maintenance fee of 10% to 20% of the purchase price is required While perpetual license software will continue to work as delivered forever, most suppliers will require that maintenance fees be kept current if the user wants updates at any time in the future In the case of annual licenses and/or leases software, unless the contract is extended and paid-up, the software will cease to work at the end of the contract period Updates during the contract period are generally included in the price, but there is some additional charge for technical support In addition to the cost of the software, training and maintaining the skills of at least one software user are not insignificant Also, unless users work regularly with the software and have a reasonable understanding of molding and mold-building processes, the results of the analysis can be disappointing and less valuable when compared with the results obtained by an operator with more experience and understanding For these reasons, it may be more practical to hire a consultant to the analysis As with any consultant, one should check the consultant’s experience and references in order to have a high degree of confidence in the results of the analysis The cost of an analysis will depend on the complexity of the part, the quality of the available electronic data, and the scope of the required analysis For a superficial flow analysis, which consists of two or © Plastics Design Library three different process-conditions iterations, you might expect to pay 10% to 20% of the price of a singlecavity mold For a complete analysis that includes recommendations on how to meet your goals (not price per iteration), the price of the analysis might even exceed the cost of a single-cavity mold However, an analysis can be made in much less time than the time required to build a mold If there are fill, shrink, or warpage problems, even building a mold might not allow the mold builder to figure out what the solutions to the problems are “Cut and try” solutions rarely find the optimum mold design Consider the cost of the product or an analysis as compared to a lawsuit for nonperformance or nondelivery of an expensive mold Anne Bernhardt of Plastics & Computer said, “One customer documented to his management that he saved over $6,000,000 on one project They didn’t care if they ever used the software again.” An analysis is easiest to justify in applications where mistakes will be costly in both time-to-market and money, and in applications where productivity is critical These include high-volume, high-precision parts where a few seconds or tenths of a second would result in significant increases in production and profit, or where a reduction in the reject rate would increase profits considerably The models provided to the person doing the analysis typically require “fixing,” and modifications even before a midplane can be defined and meshed While translator error is frequently blamed, it is well recognized that poor CAD-model quality or CAD-user inexperience is usually to blame for most model problems The time to make these modifications will add to a consultant’s fee Analysis can significantly reduce the time required to fine-tune a mold design and deliver a proven mold It can prevent building a mold that is doomed to failure This means that, while it may be difficult to justify up front, an analysis of a difficult part can save far more than the cost of the analysis The costs of the software modules range from about $600 for one of the simplest modules to over $100,000 for the complete line of products Some products list at around $300,000 Midplane modeling, and mesh and fill analysis lists in 2002 at around $15,000 This is subject to sales, discounts, negotiated prices, etc Packages are not often sold at list price Sometimes extra modules will be “thrown in for free.” In determining prices, find out if the price includes training, technical support, and maintenance fees The least expensive of TMconcept® modules is the MCO decision-support module The most expensive are the filling, shrinkage/ Ch 9: Computer-Aided Analysis 138 warpage, and cooling analysis modules The holdinganalysis module is a fairly low-cost addition to the filling analysis module, which includes the meshing/model preparation module Individual analysis charges can vary widely based on the size and complexity of the part A minimum charge is likely to be around $1000 and may range upward to over $10,000 for an analysis done by a qualified operator Currently, about 90% of analyses done are to determine if a part will fill properly Only about 10% are for cooling or shrinkage/warpage analysis The minimum cost is about $300 per iteration for a simple part by a web-based product, and that is without an experienced operator Mold Master sells a package for about $600 that is only good for runner analysis If your customer requires an analysis but you don’t intend to evaluate the analysis results, you may as well buy anyone’s “el cheapo” package and check the box that requires analysis For that matter, why the analysis at all? If the finished, poorly performing project results in a lawsuit, remember that the analysis did not cost much and you got what you paid for If it is really important, you had better buy a good product or hire a good operator Ch 9: Computer-Aided Analysis 9.11 Conclusions Plastic-injection-molding analysis software works It predicts what will happen in a mold and to the resulting molding with a good degree of accuracy It is not perfect but is far better in complex molds and parts than the best experienced mold designer’s guesses, provided that the analysis software is a quality product and the software operator is well qualified When molding difficulties are anticipated or when production volumes are in the millions, then it is fairly easy to justify the cost of analysis One significant rework of a complex mold can pay for an analysis If the analysis is done early in the design stage, avoiding rework can save a great deal of time When production is in the millions, then even a few tenths of a second saved per cycle can make a significant difference in the profit margin On the other hand, for relatively simple parts and molds, especially when the production volume is low, it is difficult to justify the cost of the analysis or the programs, and a competent mold designer can produce a satisfactory mold © Plastics Design Library