JEC COMPOSITES MAGAZINE - Issue #120 - April 2018 - 58

TECHNOLOGY reliability

lmproving composite manufacturing reliability
and yield with affordable models and simulations

elena SyerkO, chriStOphe Binetruy
xEcole Centrale de Nantes, Research Institute in Civil Engineering and Mechanics
(GeM),
UMR CNRS 6183
pavel Simacecek, SureSeSh advani
Department of Mechanical Engineering
and Center for Composite Materials,
University of Delaware

Composite manufacturing still often relies on trial-and-error approaches
to design the process, making little use of the plethora of science-based
literature, models and simulations. This paper presents an approach for the
development of easy-to-use tools with no learning curve, with lower licensing costs and can provide useful information to manufacturing engineers
or technicians, helping them improve their yield and reliability with fewer
trials. Examples of such tools are presented for liquid composite moulding
processes, but can be extended to other processes.

n the past two decades, many specialized codes were developed to
model various composite manufacturing processes. However, these
codes were not widely adopted in the industrial world because of
two reasons: (1) these simulations are not integrated within the CAD
platforms and hence need additional interfaces, making the transition
difficult; and (2) the cost is beyond the reach of many small- to medium-sized companies (SMEs) due to the code licensing fees, the learning curve associated with such simulations, and the resources and time
needed to characterize these materials because the material database is
not easily available.
The industrial world recognizes and acknowledges that the trial-and-error approaches used to develop prototypes are expensive and time consuming. Moreover, composite manufacturers inherently suffer from material and process variability, which results in lower yields than expected
during the production cycle. Hence, they would welcome an alternative
approach or solution to reduce prototype-to-production times and
improve yields despite material and process variations. At first glance,
process simulations seem to offer a commercially-available solution, but
very few companies have embraced this path to improve their productivity. In fact, it seems that most industries only resort to simulation when
their trial-and-error efforts have not resulted in a successful prototype.

As most industrial people will admit, the main hurdle to using modelling and simulation tools is the learning curve necessary to become proficient. Moreover, material data is not easily available and is time consuming to characterize and sometimes difficult to relate to the software
input. It is also difficult to convert a CAD drawing of a complex part to a
manufacturing simulation software input as they do not share the same
platform. These bottlenecks have slowed the use of simulation tools in
manufacturing.
This paper discusses how one can gradually shift from trial and error to
a full virtual simulation environment to build the path from prototypes
to products. It will focus on the widely used liquid composite moulding
process, for which commercial mould filling simulations are available [13]. This process has become more commonly used for the manufacture
of complex parts and there has been some success in demonstrating that
optimal process design can lead to a robust process despite possible flow
variations from one part to the next, producing less scrap and higher
yield [4]. This can be seen in the example presented in Figure 1.
Disruptions in the filling pattern can result from the presence of small
gaps when the preform does not hug exactly the same way around corners and inserts, causing the resin to race along these gaps and dramatically changing the flow pattern. A simulation can execute all the permutations of possible flow patterns and determine the vent locations
where the resin will arrive last for most of these scenarios. This is
illustrated in Figure 1.

Fig. 1: Liquid Injection Moulding Simulation (LIMS© software developed by the University of Delaware) can predict vent locations for successful mould filling in a complex
mould, which may trigger race tracking along edges and corners resulting in 1024 possible flow patterns [1]

jec composites magazine / N°120 April 2018

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #120 - April 2018