1 SIMULATION OF CELL GROWTH IN HIGH-PRESSURE FOAM INJECTION MOLDING Chongda Wang 1 , Vahid Shaayegan 1 , Sejin Han 2 , Franco Costa 3 , Chul B. Park 1 1 - Microcellular Plastics Manufacturing Laboratory, Department of Mechanical and Industrial Engineering, University of Toronto, 5 King’s College Road, Toronto, Ontario, Canada M5S 3G8 2 - Autodesk, Inc. 2353 North Triphammer Road, Ithaca, NY 14850, USA 3 - Autodesk, Inc. 259-261 Colchester Rd., Kilsyth, VIC. 3137, Australia Abstract In this work, the prediction of final cell size of high- pressure foam injection molded parts has been attempted. An in-situ visualization technique was used to capture real-time cell growth data from high-pressure foam injection molding experiments conducted with PS and CO2. The simulated cell growth profile was compared with experimental measurements. For the PS/CO2 system, quantitative agreement (over 80%) between predicted and measured growth profile were achieved. With the validated simulation, the effect of cooling history on final cell size and cell size distribution was investigated. It is shown that in high-pressure foam injection during which all the gate nucleated cells are dissolved, final morphology is characterized by having large cells in the center and smaller cells near the skin. Introduction Parts produced by foam injection molding (FIM) are utilized in many industries such as automotive, construction, aerospace, sports and etc. Light-weighting, geometry accuracy, short production cycles and other countless advantages bring FIM a substantial amount of attention from industry and academia. In spite of its flourish, FIM has been recognized as a highly complex unsteady process due to its cyclically interrupted plasticization and gas mixing, and the number of variables involved yet have to be controlled. As a result, when it comes to the designing of a suitable foam morphology tailored a specific application, engineers have to go through a large number of painstaking experiments and optimization to reach the desired morphology. Such iterative process is not only time, labor, and material consuming, but also inevitably increases production costs and more importantly protract time to market (TTM). One of the solutions is computer simulation and prediction. Bubble growth modeling has been studied by a number of researchers [1-9]. Amon and Denson introduced the Cell Model which describes the growth of closely spaced bubbles with limited gas concentration [1]. They assumed that each cell is surrounded by a thin shell of Newtonian fluid with limited gas concentration only from which the bubble can absorb gas molecules. Afremanesh and Advani adopted the cell model and more importantly took the viscoelasticity of the shell into account [2]. The viscoelastic fluid was represented by the quasi-linear, upper-convected Maxwell model. Moreover, they compared the growth results obtained by finding the concentration gradient via explicitly solving advection- diffusion partial derivative equation (PDE) against finding it using a fitted polynomial concentration profile. Consequently, they concluded it is necessary to solve the diffusion PDE with proper boundary conditions. Ramesh et al. applied the cell model in the simulation of batch foaming [3]. They compared the difference in predicting growth profile using between Power-law based model (Newtonian fluid) and Maxwell based model (viscoelastic fluid). Without an in-situ visualization batch foaming chamber, they quenched their sample in liquid nitrogen at different timings of bubble growth and performed SEM on the fractured surface. Their results showed that growth equations derived with viscoelastic model offer more accuracy in predicting final bubble size when compared with a pure Newtonian model. Leung and et al. also simulated the bubble growth in batch foaming process [4]. Using the Maxwell viscoelastic solution and solving the diffusion equation, more importantly, incorporating Sanchez-Lacombe equation of state (SL-EOS), they were able to not only accurately predict the growth profile but also that of bubbles nucleated at different timings. Amon and Denson also attempted to simulate the bubble growth in low-pressure FIM (LP-FIM) with the help of a visualized cavity [5]. They obtained good qualitative agreement between experimental and model- predicted results, but quantitative discrepancies were shown between them. They believed that the negligence of melt elasticity and bubble coalescence in low-pressure FIM (LP-FIM) were the main contributions to the discrepancy. Moreover, due to hardware limitations, they were unable to obtain real-time growth data. Afremanesh et al. also simulated bubble growth in a highly viscous solution in LP-FIM, furthermore, they considered the pressure variations at the different thickness of the mold cavity [6]. They reported growth profile and foam density in normalized dimensionless form. They demonstrated larger bubble size in the melt front and smaller bubbles SPE ANTEC ® Anaheim 2017 / 2525
6
Embed
Simulation of Cell Growth in High-Pressure Foam …leaders.4spe.org/spe/conferences/ANTEC2017/papers/419.pdfIN HIGH-PRESSURE FOAM INJECTION MOLDING Chongda Wang1, ... of the solutions
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
SIMULATION OF CELL GROWTH
IN HIGH-PRESSURE FOAM INJECTION MOLDING
Chongda Wang1, Vahid Shaayegan
1, Sejin Han
2, Franco Costa
3, Chul B. Park
1
1 - Microcellular Plastics Manufacturing Laboratory, Department of Mechanical and Industrial
Engineering, University of Toronto, 5 King’s College Road, Toronto, Ontario, Canada M5S 3G8
2 - Autodesk, Inc. 2353 North Triphammer Road, Ithaca, NY 14850, USA
3 - Autodesk, Inc. 259-261 Colchester Rd., Kilsyth, VIC. 3137, Australia
Abstract
In this work, the prediction of final cell size of high-
pressure foam injection molded parts has been attempted.
An in-situ visualization technique was used to capture
real-time cell growth data from high-pressure foam
injection molding experiments conducted with PS and
CO2. The simulated cell growth profile was compared
with experimental measurements. For the PS/CO2 system,
quantitative agreement (over 80%) between predicted and
measured growth profile were achieved. With the
validated simulation, the effect of cooling history on final
cell size and cell size distribution was investigated. It is
shown that in high-pressure foam injection during which
all the gate nucleated cells are dissolved, final
morphology is characterized by having large cells in the
center and smaller cells near the skin.
Introduction
Parts produced by foam injection molding (FIM) are
utilized in many industries such as automotive,
construction, aerospace, sports and etc. Light-weighting,
geometry accuracy, short production cycles and other
countless advantages bring FIM a substantial amount of
attention from industry and academia. In spite of its
flourish, FIM has been recognized as a highly complex
unsteady process due to its cyclically interrupted
plasticization and gas mixing, and the number of variables
involved yet have to be controlled. As a result, when it
comes to the designing of a suitable foam morphology
tailored a specific application, engineers have to go
through a large number of painstaking experiments and
optimization to reach the desired morphology. Such
iterative process is not only time, labor, and material
consuming, but also inevitably increases production costs
and more importantly protract time to market (TTM). One
of the solutions is computer simulation and prediction.
Bubble growth modeling has been studied by a
number of researchers [1-9]. Amon and Denson
introduced the Cell Model which describes the growth of
closely spaced bubbles with limited gas concentration [1].
They assumed that each cell is surrounded by a thin shell
of Newtonian fluid with limited gas concentration only
from which the bubble can absorb gas molecules.
Afremanesh and Advani adopted the cell model and more
importantly took the viscoelasticity of the shell into
account [2]. The viscoelastic fluid was represented by the