JEC COMPOSITES MAGAZINE - Issue #120 - April 2018 - 60

TECHNOLOGY reliability

Fig. 3: Sequential VARTM infusion

Technicians would like to know how far apart the injection lines are to
be placed and when the injection line should be opened. The spacing of
the sequential injection lines depends on parameters such as the permeability of the material and distribution media (flow media) and the panel
thickness. The question about when the next injection line should be
opened requires information about the lead length, which is the distance
between the flow front in the distribution media on top and the flow surface at the bottom. If the injection line on the top surface is opened when
the resin from the previous line reaches it, the flow at the bottom will be
far behind, resulting in a dry spot. Thus, it is necessary to wait until the
flow front has gone past the injection line location by the lead length in
the flow media before opening the sequential injection line. The spacing
between sequential injection lines and lead length information (that tells
the technicians when the injection lines should be opened) can be easily
determined by web-based tools using drop-down menus. The user simply enters the material and thickness information, and the software provides time progression of flow information in the form of a chart or table.
Automated tools
The use of simple tools is usually limited to simple and uniform geometries. Addressing more advanced and sophisticated situations requires
modelling and simulation software, proper training and a meshed part
(or geometry) model. However, the time needed for the learning curve
can be reduced as the expertise required to execute the simulations and
extract the necessary values can be automated.
The ability to wrap an automated system around process modelling
can help bridge the gap between sophisticated simulations and simple
analytic and approximate solutions. For any part for which a computer
model exists, the simulation steps could be automated (generating the
mesh from the CAD file, calculating and assigning permeabilities from
the picture of a fabric, assigning viscosity from the material datasheet,
recognizing race tracking edges from geometry, etc.).
To create such a system, several process simulations can be encapsulated
in a user interface that:
- Takes simple user input, generates additional data needed, and notes
the expected accuracy.
- Builds process models, as simple as possible for the expected level
of accuracy.
- Executes the numerical process simulation, or simulations as
needed.
- Extracts the important understandable results and delivers them to
the user.

jec composites magazine / N°120 April 2018

The user is not necessarily aware that the simulation is running, and the
interface and simulation may not actually reside on his tablet but on a
website he connects to (thus reducing costs through sharing). There is
an ongoing research interest for this encapsulation driven by the need to
make manufacturing estimates available to the part designer before the
design is complete. This should prevent the design of components that
cannot be manufactured (Figure 4), and thus shorten the development
cycle [4, 9, 10].
Such automated software can reduce the learning curve and be used on
several (three in Figure 4) increasingly accurate tiers, as the design
progresses to eliminate an increasing degree of uncertainty from the
process. The system replaces unknown data and design details with
educated estimates to provide an approximate output. Depending on
the level, the software generates the necessary manufacturing model(s),
determines what issues are to be considered, optimally executes the simulations and provides the user (designer) with only the relevant output.
Additional data can be stored for review by a process engineer. For example, to obtain simple manufacturing assistance with an existing design, any novice user can simply execute this package from Tier 2.
From the designer's perspective, these systems can be similar to outof-the-box web solutions. He can provide whatever data is available
and receive a simple output (which may just be "Cannot make within
constraints"). When properly developed and tested, this approach may
ultimately prove a great benefit not only for the design cycle in large
companies, but also for smaller companies by replacing a dedicated
manufacturing analyst.

Fig. 4: Automated process analysis system. Depending on the level of uncertainty,
the user provides the available part data and receives encapsulated output.
Software controllers convert the data, execute the simulation and extract the crucial
information from the results. The operations of the software controller(s) and simulation software are transparent to the user. Note that: (*) the full simulation data can
still be stored and reviewed, and (**) as the current implementation [4, 9, 10] is to
serve the part designer, a simple cost analysis is possible

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