Benchmark of SimuWarp for LFT injection molding

»In a collaboration project with the Robert Bosch GmbH, SimuWarp was benchmarked.«

Motivation

Injection molding with thermoplastics is accompanied by thermal strains, which induce residual stresses and part warpage. Usually, two to three mold recursions are required until dimensional accuracy is achieved. Accurate prediction of part warpage is key to digital engineering and the reduction of time- and cost-wise expensive mold recursions. However, the required prediction accuracy is not yet achieved by state-of-the-art simulation tools. Simutence addresses this issue with SimuWarp, an add-on for Abaqus for advanced warpage predictions.  

 

In a collaboration project with the Robert Bosch GmbH, SimuWarp was benchmarked. Material cards for different materials have been generated and the results of warpage predictions were correlated to experimental tests. An average prediction accuracy of 79?% was achieved, which is considered a good accuracy in the context of warpage predictions.

Key takeaways

  • Simutence provides a modular virtual process chain upon add-ons for commercial software (Abaqus and Moldflow) 
  • SimuWarp enables sophisticated material modeling for part warpage and residual stress analyses including a fully automatized model setup 
  • Accurate warpage prediction enables digital design engineering and a reduction of required mold recursions. 

Integration into a modular CAE chain

Simutence add-ons follow a modular approach, to enable the usage in existing software infrastructures.  

 

In the project with Robert Bosch, SimuFill is adopted to model the crystallization kinetics during injection molding in Moldflow. Subsequently, the results relevant to warpage analysis are exported and transformed into a VTK file, which is used by Simutence as a neutral data exchange format.  

 

Finally, SimuWarp is used for warpage analysis in Moldflow. First, the model is set up fully automatized, and homogenized material cards to consider the local fiber orientation are generated. Second, the analysis is run in Abaqus/Standard in combination with the material modeling approaches supplied through SimuWarp. 

Integration into a modular CAE chain

Materials characterization

Accurate materials characterization is the essential basis for accurate predictions. In the project with Robert Bosch, crystallization kinetics and viscoelasticity were characterized among others for a PBT reinforced with 30?wt.% glass fibers. In contrast, data from the Moldflow material library was used for the pvT, heat capacity, and heat conductivity properties.

Crystallization kinetics

Usually, semi-crystalline thermoplastics are chosen for engineering applications due to their improved mechanical properties. Crystallization defines the transition from the molten to the solid material state for semi-crystalline thermoplastics. During crystallization, thermal strains due to shrinkage and mechanical properties evolve. Thus, accurate modeling of crystallization kinetics is essential to accurately predict thermal strains as well as the resulting residual stresses and warpage.  

 

Differential Scanning Calorimetry (DSC) was used to characterize the crystallization kinetics. Subsequently, parameters for the Nakamura-Ziabicki model were identified, to model crystallization kinetics with SimuFill and SimuWarp as a function of the cooling rate. 

Crystallization kinetics

Fitting result using the Nakamura-Ziabicki model for non-isothermal crystallization (solid: experimental, dashed: Nakamura-Ziabicki).

Viscoelasticity

A large portion of the processing of thermoplastics involves temperatures above the glass transition temperature. Here, relaxation times are short and thus stress relaxation plays a major role in warpage analyses. SimuWarp enables the modeling of stress relaxation through viscoelastic material modeling. 

 

Dynamic Mechanical Analysis (DMA) was used to characterize the viscoelastic mechanical behavior. Time Temperature Superposition (TTS) was used to generate a master curve for the rheological simple material behavior. Based on this, stress relaxation is modeled over a wide range of temperatures and relaxation times. 

 

Viscoelasticity

Master curve characterized through Dynamic Mechanical Analysis (DMA) and generated through Time Temperature Superposition (TTS). 

pvT

Thermal strains are induced during processing by shrinkage and temperature changes. SimuWarp uses pvT-based modeling of thermal strains using the 2-domain Tait model.  

 

In the scope of this project, 2-domain Tait parameterization from the Moldflow material library was used. Otherwise, pvT characterization could have been conducted. 

LFT specific volume

Specific volume as a function of hydrostatic pressure and temperature.

Input from molding simulation with SimuFill

 

An injection molding simulation in Moldflow in combination with our add-on SimuFill is conducted for filling and packing. SimuFill enhances the modeling capabilities of Moldflow to predict crystallization kinetics.  

 

Based on the injection molding analysis with SimuFill, relevant state variables, such as the local fiber orientation, temperature, or crystallization are predicted. Moreover, the time and temperature at zero pressure are predicted, which is required to address the shrinkage compensation due to the packing pressure in the subsequent warpage analysis. These state variables serve as input for the warpage simulation in Abaqus. 

Injection molding simulation in Moldflow using SimuFill. 

 

Advanced warpage simulation with SimuWarp

 

SimuWarp enables to predict warpage and residual stresses with Abaqus, resulting in the following benefits: 

  • Sophisticated material modeling
    (viscoelasticity, crystallization kinetics) 
  • Accurate modeling of the cooling and boundary conditions
    (part ejection, gravity, assembly) 
  • Consideration of local reinforcements and hybrids 
  • Utilities that facilitate and automate the setup of the simulation model 

 

Based on this, SimuWarp improves the modeling capabilities for part warpage and residual stresses compared to conventional tools. Moreover, it includes utilities that facilitate and automate the setup of the simulation model. 

Warpage simulation in Abaqus using SimuWarp

Validation of warpage prediction 


In the scope of this project, molded parts have been scanned and characteristic distances have been measured. The correlation of measured and virtual deformation provides the prediction accuracy. 

 

An average prediction accuracy of 77 % is observed, which is a high degree of agreement in the context of warpage analysis. 

Any Questions?

Do not hesitate to get in contact with us if you have any questions or if you are interested in a collaboration. We are looking forward to receiving your request!

References

  • Lückenfülli 1
  • und 2