Spring’96 Volume 2.1

11 programs, 3 service levels

Datapoint Testing Services now provides analysis-ready material data for eleven design programs, spanning the areas of mold analysis, extrusion and structural analysis. For maximum convenience, three levels of service are available. The DATAPOINT PRIORITY service is geared toward clients in emergency situations with data delivered within 48 hours, frequently on the same day. The DATAPOINT STANDARD service, still the fastest in the industry, provides convenient one week delivery on all tests. Clients who have the time are encouraged to use DATAPOINT ECONOMY service For a slightly extended delivery, significant savings can be achieved.

New developments in CAE: improving solution accuracy

Mold analysis programs continue to make strides in improving the accuracy of their predictions, by accounting for more and more of the complex phenomena which occur during the molding process In order to characterize the material’s response to these phenomena and provide proper inputs for the simulations. new schemes for material testing have been developed.

Using visco-elastic data for shrink-warp modeling

The solidification of polymers following injection molding tends to produce residual stresses due to flow and thermal effects Because of the visco-elastic nature of polymers the stresses will tend to relax over a period of time, with the rate of decay decreasing sharply with temperature until the onset of solidification phenomena such as a glass transition or crystallization At this point any stresses still remaining in the part will be frozen in High levels of residual stress and localized regions of stress are causes for poor mechanical strength and warpage in the part. Because of the nature of the injection molding process, core regions will typically have longer times at higher temperatures, resulting in a lower level of residual stresses. On the other hand the skin regions cool much faster and are more susceptible to frozen-in stresses To account for these phenomena effectively, a model for the stress relaxation of the material should provide a proper representation of the manner in which the residual stresses in the material will relax.

Maxwell model (spring/dashpot) used to represent visco-elastic stress relaxation in plastics

Both C-MOLD and 1-DEAS provide viscoelastic capability in order to model this phenomenon. To generate data for 1-DEAS stress relaxation data are obtained from dynamic mechanical analysis. in which specimens are tested in torsion over a range of frequencies and temperatures below the melting temperature of the polymer. The resulting data are reduced using time- temperature superposition, and the WLF parameters are obtained The data are transformed into a stress relaxation curve which, along with the flow analysis properties and PVT data, constitutes the material data needed to run such an analysis. According to SDRC technical support, visco elastic simulations are essential in order to make accurate shrinkage and warpage predictions

Datapoint Testing Services is part of the I-DEAS Solutions Network and provides viscoelastic and other characterizations for l-DLAS programs.

Improving pressure predictions in runners and gates

In flows through restrictions such as gates and in runner systems with rapid reduction in diameter, additional pressure drop is experienced over that which occurs due to flow through the cross section. This term is not accounted for in conventional flow simulations, with the result that the pressure loss across gates and flow restrictions is underpredicted when compared to experimental data. Now, both C-MOLD and Moldflow have introduced the capability to model pressure drops in the runner system and gate.

Moldflow’s approach uses extensional viscosity, which measures a fluid’s ability to resist extension. Based on a scheme originally proposed by F. N. Cogswell and refined by D. M. Binding, data from a capillary rheometer fitted with an orifice die are used to measure extensional viscosity, which can then be used to model this phenomenon. This enhancement is now available for MF/FLOW and its derivative products (MF/OPTIM, MF/GAS, and the fiber option).

Smooth flow with no vortices (a) and with corner vortices (b).

The C-MOLD filling program performs a juncture-loss analysis using parameters obtained from a Bagley correction of capillary rheometer data. Here, Bagley fit parameters are used to calculate the end pressure loss contribution to the overall pressure drop. Juncture loss calculations activate when the ratio of diameters of two adjacent runner elements is greater than 2:1. C-MOLD support staff say that pressure predictions can be improved 7 to 20%, depending on the material and geometry In situations where the Bagley parameters are not provided, the program uses an internal model to estimate the end pressure loss.

Properties needed for using these schemes can be generated by Datapoint Testing Services and by the respective software vendors.

Non-linear material models for MF-STRESS

Polymers exhibit a non-linear stress/strain relationship when subject to large scale deformations. Except for brittle polymers, non-linear deformations tend to be the rule rather than the exception when evaluating the performance of plastic parts. A typical unfilled thermoplastic, which has a relatively low elastic modulus, may undergo more than an order of magnitude more strain than a metal before exhibiting plastic deformation. Further, the elastic deformations are nonlinear and the strain is recoverable, though over a period of time. The new non-linear material model in Moldflow’s MF-STRESS accounts for this kind of behavior. The primary advantage here is that the residual stress states and anisotropic orientations of the part, as developed during the molding simulation, can be used as initial conditions for finite element analysis.

Four sets of stress strain data are needed for this model.
Tensile stress strain curves parallel to flow
Tensile stress strain curves
perpendicular to flow
Compressive stress strain curves parallel to flow
Compressive stress strain curves perpendicular to flow

Datapoint Testing Services performs these characterizations for MF-Stress and other structural analysis programs.

TMconcept adopts a hybrid approach to using compressibility data

The data for faHold and faShape is based on pvT data integrated with DSC and quenching measurements. According to Anne Bernhardt at Plastics & Computer, the characterization takes into account the degree of crystallinity and hence accurately predicts shrinkage. A database of generic data which has been treated in this manner is available for almost all categories of thermoplastic resins.

Hubert Lobe, Data point Testing Services.

Tips for Practitioners premieres

A pilot Tips for Practitioners symposium at this years ANTEC will focus on the rapidly advancing fields of mold analysis, structural analysis and process simulation. Titled “Design Analysis: Tools for Concurrent Engineering”, the symposium will feature prominent practitioners in this field. Each will present a 15 minute talk in their area of expertise. The presentations will, seek to provide practical tips on the use of design analysis; talk about interpretation of analysis results; provide deeper insights to help improve the understanding of these tools; and present new ways of using simulation programs.

The symposium will be held at ANTEC, Thursday, May 9th, 2:00 - 4:30 pm, in Indianapolis.

When correct material property data is unavailable...

The importance of correct material data inputs for injection molding engineering analyst’s of derqna cannot be over stressed. As with any analysis the quality of the results can only be as good as the quality of the inputs. regardless of what analysis technique or software is used The degree of criticality of the material data inputs is dependent upon both the predictions the analyst is looking to extract from the results and the complexity of the part design The answer to whether grade specific data is needed for good results s likely to be different for weld line predictions of a nominal waf thickness design molded in ABS. CS opposed to maximum warpage predictions for a variable wal thickness part molded si an engineering alloy

Analysis with a material suhsttntion could be beneficial when a design optimization is being performed such as cooling circuit design. Cooling circurtrv analysis to deliver un form heat extraction is primarily a function of geometry (part and cooling circuits) so an optimized design should work for any material even frorr a different plastic family. if accuracy of the cycle prediction and other specifics are not needed Another instance where material suhstitjtiofl might work would be in comparisons of gatirig scenarios

Polymer behavior is much more complex than might be expected and there are many situations than can produce unexpectedly different behavior in similar polymers Thickness variations in a part is one source of possible error. The material’s thermal diffusivity (thermal / conductivity specific * heat melt density) and viscosity am the primary factors that will determine if hesitation of flow might occur when the polymer flows from a thick to thin area if a polymer has a moderately higher thermal diffusivty along with higher viscosity than the second, hesitation might occur n the first material and not tire second under a given set of processing conditions This is triggered by the interaction of the decreased flow channel thickness from a faster / thicker frozen layer and the higher fill pressures due to the higher viscosity Two similar polymers might also have different sensitivities of viscosity relative to shear rate. This might produce similar molding characteristics Ipresaure weld iines etc ) at fast injection rates, but when molded at slower injection rates. the two polymers might exhibit significant differences in behavior.

Analysis past simple fill pattern prediction is much more sensitive to material data accuracy Shrinkage and resulting wamoaqr. predictions are a result of alt of the material characteristics thermal characteristics rheology, PVT and other shrinkage specific behavior and mechanical properties A recent arratyss performed for a customer reemphasized this point 1 he part was to he molded in a general purpose. unfilled homopolymer PP. Representative PP data was selected for the initial filling, cooling and sdrinkage!warpage analysis gate location and processing were optimized for that material. However, part flatness was just not adequate to warrant pursuing this as a new project Unexpectedly. when data from another PP was analyzed under the same conditions. The predictions were that the part would be significantly flatter and well within the design requirements. Both polymers were nucleated, unfilled homopolymers with the same quoted melt flow rate The interaction of all of the variations in material properties between the two polymers produced significant differences in sensitivity of shrinkage to polymer orientation and packing pressure.

It is unreasonable to expect that grade specific data will be available for all injection molding materials. There will always be the question "What material data is available that I can substitute or should I not try to substitute?” I hope this brief article has helped clarify some of the factors involved.

For further information on the subject of analysis sensitivity to material data, see ‘The importance of Accurate Material Data in the Design and Manufacture of Injection Molded Parts’ presented at the 1992 SME Mold Filling and Cooling Analysis Clinic in Chicago This presentation was also summarized in ‘Making the Most of CAE Software’ in the September 1993 issue of Plastics Design Forum

Mr Robert Sherman is Engineering Manager at Bluegi ass Plastics Engineering. Inc.


	Spring ’96 Volume 2.1
	Articles 11 programs, 3 service levels New developments in CAE: improving solution accuracy Tips for Practitioners premieres When correct material property data is unavailable...
	11 programs, 3 service levels Datapoint Testing Services now provides analysis-ready material data for eleven design programs, spanning the areas of mold analysis, extrusion and structural analysis. For maximum convenience, three levels of service are available. The DATAPOINT PRIORITY service is geared toward clients in emergency situations with data delivered within 48 hours, frequently on the same day. The DATAPOINT STANDARD service, still the fastest in the industry, provides convenient one week delivery on all tests. Clients who have the time are encouraged to use DATAPOINT ECONOMY service For a slightly extended delivery, significant savings can be achieved. New developments in CAE: improving solution accuracy Mold analysis programs continue to make strides in improving the accuracy of their predictions, by accounting for more and more of the complex phenomena which occur during the molding process In order to characterize the material’s response to these phenomena and provide proper inputs for the simulations. new schemes for material testing have been developed. Using visco-elastic data for shrink-warp modeling The solidification of polymers following injection molding tends to produce residual stresses due to flow and thermal effects Because of the visco-elastic nature of polymers the stresses will tend to relax over a period of time, with the rate of decay decreasing sharply with temperature until the onset of solidification phenomena such as a glass transition or crystallization At this point any stresses still remaining in the part will be frozen in High levels of residual stress and localized regions of stress are causes for poor mechanical strength and warpage in the part. Because of the nature of the injection molding process, core regions will typically have longer times at higher temperatures, resulting in a lower level of residual stresses. On the other hand the skin regions cool much faster and are more susceptible to frozen-in stresses To account for these phenomena effectively, a model for the stress relaxation of the material should provide a proper representation of the manner in which the residual stresses in the material will relax. Maxwell model (spring/dashpot) used to represent visco-elastic stress relaxation in plastics Both C-MOLD and 1-DEAS provide viscoelastic capability in order to model this phenomenon. To generate data for 1-DEAS stress relaxation data are obtained from dynamic mechanical analysis. in which specimens are tested in torsion over a range of frequencies and temperatures below the melting temperature of the polymer. The resulting data are reduced using time- temperature superposition, and the WLF parameters are obtained The data are transformed into a stress relaxation curve which, along with the flow analysis properties and PVT data, constitutes the material data needed to run such an analysis. According to SDRC technical support, visco elastic simulations are essential in order to make accurate shrinkage and warpage predictions Datapoint Testing Services is part of the I-DEAS Solutions Network and provides viscoelastic and other characterizations for l-DLAS programs. Improving pressure predictions in runners and gates In flows through restrictions such as gates and in runner systems with rapid reduction in diameter, additional pressure drop is experienced over that which occurs due to flow through the cross section. This term is not accounted for in conventional flow simulations, with the result that the pressure loss across gates and flow restrictions is underpredicted when compared to experimental data. Now, both C-MOLD and Moldflow have introduced the capability to model pressure drops in the runner system and gate. Moldflow’s approach uses extensional viscosity, which measures a fluid’s ability to resist extension. Based on a scheme originally proposed by F. N. Cogswell and refined by D. M. Binding, data from a capillary rheometer fitted with an orifice die are used to measure extensional viscosity, which can then be used to model this phenomenon. This enhancement is now available for MF/FLOW and its derivative products (MF/OPTIM, MF/GAS, and the fiber option). Smooth flow with no vortices (a) and with corner vortices (b). The C-MOLD filling program performs a juncture-loss analysis using parameters obtained from a Bagley correction of capillary rheometer data. Here, Bagley fit parameters are used to calculate the end pressure loss contribution to the overall pressure drop. Juncture loss calculations activate when the ratio of diameters of two adjacent runner elements is greater than 2:1. C-MOLD support staff say that pressure predictions can be improved 7 to 20%, depending on the material and geometry In situations where the Bagley parameters are not provided, the program uses an internal model to estimate the end pressure loss. Properties needed for using these schemes can be generated by Datapoint Testing Services and by the respective software vendors. Non-linear material models for MF-STRESS Polymers exhibit a non-linear stress/strain relationship when subject to large scale deformations. Except for brittle polymers, non-linear deformations tend to be the rule rather than the exception when evaluating the performance of plastic parts. A typical unfilled thermoplastic, which has a relatively low elastic modulus, may undergo more than an order of magnitude more strain than a metal before exhibiting plastic deformation. Further, the elastic deformations are nonlinear and the strain is recoverable, though over a period of time. The new non-linear material model in Moldflow’s MF-STRESS accounts for this kind of behavior. The primary advantage here is that the residual stress states and anisotropic orientations of the part, as developed during the molding simulation, can be used as initial conditions for finite element analysis. Four sets of stress strain data are needed for this model. Tensile stress strain curves parallel to flow Tensile stress strain curves perpendicular to flow Compressive stress strain curves parallel to flow Compressive stress strain curves perpendicular to flow Datapoint Testing Services performs these characterizations for MF-Stress and other structural analysis programs. TMconcept adopts a hybrid approach to using compressibility data The data for faHold and faShape is based on pvT data integrated with DSC and quenching measurements. According to Anne Bernhardt at Plastics & Computer, the characterization takes into account the degree of crystallinity and hence accurately predicts shrinkage. A database of generic data which has been treated in this manner is available for almost all categories of thermoplastic resins. Hubert Lobe, Data point Testing Services. Tips for Practitioners premieres A pilot Tips for Practitioners symposium at this years ANTEC will focus on the rapidly advancing fields of mold analysis, structural analysis and process simulation. Titled “Design Analysis: Tools for Concurrent Engineering”, the symposium will feature prominent practitioners in this field. Each will present a 15 minute talk in their area of expertise. The presentations will, seek to provide practical tips on the use of design analysis; talk about interpretation of analysis results; provide deeper insights to help improve the understanding of these tools; and present new ways of using simulation programs. The symposium will be held at ANTEC, Thursday, May 9th, 2:00 - 4:30 pm, in Indianapolis. When correct material property data is unavailable... The importance of correct material data inputs for injection molding engineering analyst’s of derqna cannot be over stressed. As with any analysis the quality of the results can only be as good as the quality of the inputs. regardless of what analysis technique or software is used The degree of criticality of the material data inputs is dependent upon both the predictions the analyst is looking to extract from the results and the complexity of the part design The answer to whether grade specific data is needed for good results s likely to be different for weld line predictions of a nominal waf thickness design molded in ABS. CS opposed to maximum warpage predictions for a variable wal thickness part molded si an engineering alloy Analysis with a material suhsttntion could be beneficial when a design optimization is being performed such as cooling circuit design. Cooling circurtrv analysis to deliver un form heat extraction is primarily a function of geometry (part and cooling circuits) so an optimized design should work for any material even frorr a different plastic family. if accuracy of the cycle prediction and other specifics are not needed Another instance where material suhstitjtiofl might work would be in comparisons of gatirig scenarios Polymer behavior is much more complex than might be expected and there are many situations than can produce unexpectedly different behavior in similar polymers Thickness variations in a part is one source of possible error. The material’s thermal diffusivity (thermal / conductivity specific * heat melt density) and viscosity am the primary factors that will determine if hesitation of flow might occur when the polymer flows from a thick to thin area if a polymer has a moderately higher thermal diffusivty along with higher viscosity than the second, hesitation might occur n the first material and not tire second under a given set of processing conditions This is triggered by the interaction of the decreased flow channel thickness from a faster / thicker frozen layer and the higher fill pressures due to the higher viscosity Two similar polymers might also have different sensitivities of viscosity relative to shear rate. This might produce similar molding characteristics Ipresaure weld iines etc ) at fast injection rates, but when molded at slower injection rates. the two polymers might exhibit significant differences in behavior. Analysis past simple fill pattern prediction is much more sensitive to material data accuracy Shrinkage and resulting wamoaqr. predictions are a result of alt of the material characteristics thermal characteristics rheology, PVT and other shrinkage specific behavior and mechanical properties A recent arratyss performed for a customer reemphasized this point 1 he part was to he molded in a general purpose. unfilled homopolymer PP. Representative PP data was selected for the initial filling, cooling and sdrinkage!warpage analysis gate location and processing were optimized for that material. However, part flatness was just not adequate to warrant pursuing this as a new project Unexpectedly. when data from another PP was analyzed under the same conditions. The predictions were that the part would be significantly flatter and well within the design requirements. Both polymers were nucleated, unfilled homopolymers with the same quoted melt flow rate The interaction of all of the variations in material properties between the two polymers produced significant differences in sensitivity of shrinkage to polymer orientation and packing pressure. It is unreasonable to expect that grade specific data will be available for all injection molding materials. There will always be the question "What material data is available that I can substitute or should I not try to substitute?” I hope this brief article has helped clarify some of the factors involved. For further information on the subject of analysis sensitivity to material data, see ‘The importance of Accurate Material Data in the Design and Manufacture of Injection Molded Parts’ presented at the 1992 SME Mold Filling and Cooling Analysis Clinic in Chicago This presentation was also summarized in ‘Making the Most of CAE Software’ in the September 1993 issue of Plastics Design Forum Mr Robert Sherman is Engineering Manager at Bluegi ass Plastics Engineering. Inc.