“Frontal polymerisation doesn’t use an autoclave at all, so it doesn’t require that huge upfront investment,” said Bliss Professor Philippe Geubelle in the Department of Aerospace Engineering at the U of I. “It’s a chemical reaction sustained by the release of heat as the front propagates. It can save a lot of energy and it generates much less carbon dioxide, so that’s an environmental benefit.”

Geubelle said they began comparing the two methods by looking at the thermo-chemical equations in order to model the two polymerisation processes. In that way, they could compare the methods for a variety of composite materials, and particularly, the time duration each method takes to manufacture the same part.

“The key contribution from the theoretical point of view is we’ve rewritten the reaction-diffusion equations to extract the two most important non-dimensional parameters,” Geubelle said. “Using just these two parameters allowed us to look at a wide range of chemical parameters, such as the activation energy and the heat of reaction, and at the impact of the initial temperature of the resin.”

Geubelle said this method helped to compare the composite manufacturing processes based on bulk and frontal polymerisation in terms of the time it takes to manufacture a part. The study found that there were instances when one or the other was faster.

“Imagine you want to make something that is one meter long. Frontal polymerisation will be able to do complete the task before bulk polymerisation starts to kick in,” Geubelle said. “On the other hand, if you want to make something that is 10 meters long, then bulk polymerisation may actually take place before the front reaches the other end of the part. It’s the competition between these two processes that we analysed in this study.”

He went on to say there are several ways to speed up the process for frontal polymerization: start the front at both ends so it goes twice as fast, or heat it from the bottom by using a heated panel beneath it. “That process is so fast, we refer to it as flash curing,” Geubelle said, “but it does use more energy than for a single front.”

Manufacturing composite parts using frontal polymerization instead of bulk polymerization has a lot of advantages.

“With frontal polymerisation, you don’t need the large capital investment of the autoclave, making it a very attractive option,” Geubelle said. “The time it takes to cure a composite part is also much shorter and the environmental impact is substantially reduced.”

The study, “Frontal vs. bulk polymerization of fibre-reinforced polymer-matrix composites,” was written by S. Vyas, X. Zhang, E. Goli, and P. H. Geubelle. It is published in Composites Science and Technology.

The post Analysing the pros and cons of two composite manufacturing methods appeared first on Composites Today.

Composite laminates used in Aerospace applications are typically made of layers of carbon fibres with varying orientations embedded in epoxy. For example, the composite laminate can be composed of a carbon/epoxy layer with the fibres oriented in the 90-degree direction sandwiched between two 0-degree plies. The fibres are each about seven microns in diameter, or about one-seventh of the thickness of a human hair.

“We know from experiments that cracks propagate transversely across the 90-degree plane, then stop when they reach the interfaces with the 0-degree plies. So we developed a method that allows us to simulate hundreds of fibres in a realistic system and study how the failure response is affected if we change the location of a single fibre or of many fibres, or the strength of the interface,” said Philippe Geubelle, a professor in the Department of Aerospace Engineering.

In this new method, optical micrographs are taken of the 90-degree ply and the location of all of the fibers are extracted to construct a realistic computational model of the ply. Similar studies have been limited to tens of fibers.

“With the special finite element method we have developed to simulate the transverse cracking of the 90-degree ply, we can simulate hundreds of fibres,” Geubelle said. “The most we’ve done so far is close to 3,000 fibres.”

“Because the crack propagates primarily along the fibre-matrix interfaces, our model emphasizes the cohesive failure of these interfaces,” he said. “In addition, we have developed the ability to extract efficiently the sensitivity of the failure event with respect to the properties of the microstructure including the location and size of the fibres, and the failure properties of the fibre-matrix interfaces. We can also compute the sensitivity of the failure event with respect to the parameters (average, standard deviation, etc.) that define the distribution of these microstructural parameters.”

“Of course, you could get these sensitivities experimentally, with every conceivable variation, to see what the effect is on the failure event,” Geubelle said. “To do this numerically is much more efficient.”

Of course, you could get these sensitivities experimentally, with every conceivable variation, to see what the effect is on the failure event,” Geubelle said. “To do this numerically is much more efficient.

The study, “Transverse Failure of Unidirectional Composites: Sensitivity to Interfacial Properties,” written by  S. Zacek, D. Brandyberry, A. Klepacki, C. Montgomery, M. Shakiba, M. Rossol, A. Naja, X. Zhang, N. Sottos, P. Geubelle, C. Przybyla, and G. Jefferson appears in Integrated Computational Materials Engineering.

The post Simulating Failure to Find the Strength in Composites appeared first on Composites Today.