Design of Cure Cycles for Thick Section Composites

1. Design a cure cycle for a composite plate (comprised of stacks of woven glass fabric impregnated with a thermosetting vinylester resin) that yields good part quality in minimum processing time.

2. Evaluate the performance of the cure cycle by fabricating the plate and
   • monitoring temperatures within the plate during curing stage of the experiment;
   • inspecting the cured plate for voids, delaminations, poor compaction.

Existing cure simulation computer programs and literature values for input parameters will be used to predict temperature and extent of reaction profiles through the thickness of the composite plate for various temperature cycles of the heat press. A preliminary temperature cycle will be selected which minimizes the peak temperature observed in the center of the composite, ensures the composite cures inside/out once the gel point is reached, and maintains a curing time less than 1.5 hours. The sensitivity of the above criteria will be investigated with respect to variations in key input parameters. An experimental characterization program will be designed to obtain values of the key input parameters for the current batch of materials. If necessary, the cure cycle will be modified to reflect any batch-to-batch variation of the current material from previously reported values.

During fabrication, thermocouples will be embedded at appropriate locations within the composite to monitor the temperature at various locations during cure. Variations from model predictions will be used to evaluate the design of the cure cycle

The cured part will be inspected by microscopy to evaluate the quality and uniformity of the compaction of the plies; the content of polymer and void in the finished part will be determined.

Currently, the dominant form of composite materials utilize thermosetting polymeric materials as the matrix phase. In the presence of a catalyst, heat, radiation and/or pressure, thermosetting resins solidify through an irreversible exothermic chemical reaction or "cure". Prior to cure, the polymer is a viscous fluid that can be made to flow under pressure. As the cure reactions proceed, the molecules react to form a covalently bonded three-dimensional network. The increasing molecular weight is accompanied by an increase in viscosity. At the gel-point, a loose three-dimensional network pervades the system, the polymer exhibits the behavior of a gel, and flow ceases. However, the reactions continue and form a tightly cross-linked structure with the characteristics of a glassy solid.

The resin transfer molding (RTM) process is widely used to fabricate composite materials. Dry reinforcing fibers are impregnated by a low viscosity thermosetting resin, which, upon curing, provides integrity to the assembly. The process is begun by placing the dry fiber sheets into the mold cavity. Thermocouples are placed within the mold at this time to later monitor the temperatures within the part during the curing cycle. Once the fibers are placed and the mold is closed, the resin is injected into the mold. This is achieved by applying pressure (via compressed air) to the surface of a resin reservoir, which forces the resin into the mold and through the fiber preform. The injection process is continued until no air bubbles are seen in the exiting resin.

Once the mold is completely filled, the curing cycle of the resin can begin. The water heater is set to a temperature that was deemed adequate from the simulation results. Once the resin in the center of the composite begins to gel (as seen by a large temperature increase, or "exotherm"), it may be desirable to increase the water temperature to reduce the temperature difference between the center and the surface of the composite and speed up the curing process. Although this should be verified by the simulation, realize that the simulated results may not accurately place the exotherms in time, due to inaccuracies in the model parameters. Once the entire composite is thought to have reached over 90% cured, the mold can be separated and the part removed. The part is then placed in an oven at 120 °C to complete the cure.

A minimal successful cure cycle is defined in terms of several non-mechanical criteria:

• all of the polymer must have achieved a minimum acceptable degree of cure;
• the formation of voids must be prevented such that the number and size of all voids are below an acceptable limit;
• the temperature must remain below some critical value at which damage to the structure may occur through polymer degradation;
• the part must be cured uniformly such that large gradients in degree of conversion or temperature are avoided -- the presence of such gradients will result in residual stresses which detract from the performance of the part.

The goal of the cure cycle designer is to meet these criteria by properly selecting the temperature cycle applied to the laminate to allow complete, uniform, controlled cure while completing the cure process in as short a period of time as possible to reduce the cost of producing the part. A major problem encountered in curing thick section laminates is the control of the reaction exotherm and heat transfer in order to achieve uniform cure and minimum residual stresses. The problem is aggravated by the poor and non-isotropic thermal conductivity of the assembly.

The traditional practice in cure cycle design has been to cure several parts by different trial cycles until a cycle was obtained which resulted in a satisfactory product. This procedure is lengthy, costly, and inefficient; it does not insure optimal cure cycles. Recently, efforts have been directed at developing models which simulate the cure process in order to facilitate the design of cure cycles. These models can also serve as the basis of a monitoring system for on-line control of the cure process.

The cure process is characterized by three physical-chemical processes:

• mass transfer with chemical reactions;
• heat transfer with internal heat generation;
• the fluid dynamics of the flow of the polymer.

The chemical and physical information required for the current selection of materials can be grouped into three categories.

Reaction Characteristics: The pertinent cure reactions must be identified. Knowledge is required of the reaction kinetics and reaction activation energy as determined from initial reaction rates at various temperatures in the appropriate range. In some cases the role of diffusion should be taken into account in the later states of the cure as the mobility of the reacting species is diminished. The heat of reaction and the rate of heat generation must be available. Fourier Transform Infrared Spectroscopy (FTIR) and/or Differential Scanning Calorimetry (DSC) provide useful characterization tools for obtaining this data.

Thermal Characteristics: Descriptors such as thermal conductivity, specific heat and density are required. These descriptors for the laminate are frequently estimated by employing "Combining Rules" that predict laminate properties in terms of the stacking geometry and the volume fraction concentration and properties of the fiber and polymer components.

Chemo-Rheological Characteristics: The viscosity of the polymer as a function of temperature and extent of cure should be available. In addition, the gelation time and the dependence of the "Glass Transition Temperature" (Tg) on extent of cure should be known.

Prior to evaluating mechanical performance, the part should be inspected to identify flaws that detract from the performance. Local areas in which the plies did not bond can be identified by nondestructive inspection with ultrasonic techniques. An evaluation of the uniformity of the part usually requires destructive sectioning and microscopic examinations to determine (i) variations in the final thickness of the individual plies throughout the thickness of the part, and (ii) the number and size distribution of small voids. Interpretation of mechanical testing data requires knowledge of the weight or volume fraction of fiber and resin in the finished part. This data can be obtained by "ashing" or acid digestion to remove the polymer from a representative specimen of known mass. If the part is essentially free of voids, the concentration of fiber in the finished part may be determined from density measurements obtained with a density gradient column.

The following suggestions are offered as guidelines for accomplishing the objectives of this assignment.

A Prior to the scheduled pre-lab interview, the student team should:

• Gain familiarity with the language and concepts associated with the project.
• Identify resources and collect the required technical information.
• Schedule an appointment with the instructor and/or teaching assistant to gain clarification of issues and procedures.
• Exercise the cure simulations to establish sensitivities to input parameters.

• Justification for the selection of the particular characterization experiments.
• Specification of the assignment and responsibilities of each team member.
• A realistic schedule of activities consistent with delivering a timely final report and recommendations.

• Key results from the sensitivity analysis of the cure simulation.
• Key results from the selected characterization experiments.
• Drafts of Figures, Tables, and Schematics that will appear in the final report.
• An outline of the final report.

• A brief statement of the problem which emphasizes the major issues.
• A succinct review of the approach and concepts used in establishing the model relationships used in the simulation. The key simplifying assumptions should be identified.
• A summary of all input parameters used in the simulation. Input parameters obtained from new characterizations should be identified.
• Appropriate plots comparing predicted and measured temperature profiles.
• Appropriate photographs and/or plots describing the quality of the product.
• An evaluation of the proposed design of the cure cycle.
• An assessment of the limitations of the current Thermokinetic and Heat Transfer models.
• Recommendations for future work.

Papers & Reports:

"Composite Materials" J. K. Lees, A. K. Dingra, and R. L. McCullough, Ullmann's Encyclopedia of Industrial Chemistry, A7, (Fifth Edition) pg. 376-410 (1986).

"Processing Science of Thick-Section Composites," D. E. Hanks, M. C. Lee, R. C. Young, and Y. A. Tajima, SAMPE Quarterly, 19, No. 2 (January) pg. 19-28 (1988). A Case Study.

"Reinforced Plastics: Theory and Practice," M. W. Gaylord, Koppers Co. publication, pg. 27-38 (1969)

"Resin Transfer Molding," L. Fong and S. G. Advani, Center for Composite Materials, University of Delaware, CCM Report 95-09, April 1995.

"Cure Behavior of Unsaturated Polyester Resin Composites," D. C. Adams, Senior Thesis, Department of Chemical Engineering, University of Delaware, CCM Report 88-16, (1988).

ÒProcess Induced Stress and Deformation in Thick-Section Thermosetting Composites,Ó Travis A. Bogetti, Ph.D. Dissertation, Department of Mechanical Engineering, University of Delaware, CCM Report 89-32, Chapters 2 and 3, pp. 14-114.

Computer Programs:

Characterization Equipment:

Fabrication Equipment:

Back to Composites Lab Page