Daher CARAC TP project advances thermoplastic composites certification approach

Thermoplastic composites enable lightweight, lower-emissions aircraft with the potential for shorter cycle times and lower cost. “Daher is manufacturing both thermoset and thermoplastic composite parts mostly of medium complexity and small-to-medium dimensions,” says Martin Denize, composite R&T project manager at Daher. “Now, we want to produce more complex parts with larger dimensions for applications on wings and engines because this is what our customers are asking for. For this, we need to understand how the thermoplastic materials behave to go from development to the certification of those parts.”

This is the core of Daher’s CARAC TP project — characterization of thermoplastics — to climb the testing pyramid and achieve certified materials in challenging service environments, including temperatures above the glass transition temperature (T_g) of the thermoplastic composite (TPC) materials. “We believe high-performance TPCs (e.g., PEEK, PEKK and LM-PAEK) don’t have the same degradation issue as epoxy composites under high-temp and hot/wet conditions,” says Stephanie Patel, materials department manager at Daher. “Commonly, organic composite materials are used 28°C under wet T_g. Part of CARAC TP is to make sure there are no showstopper issues for these TPCs near T_g, but we also want to compare them to thermoset materials.”

This is important, adds Denize, “because we are also developing parts with new fast-curing and out-of-autoclave [OOA] thermoset resins, and we want to know which material is best for each application.”

“We want to have the knowledge to look at all materials and develop the part from beginning to end … pick the right material, the right process with the right parameters and know how it will perform in the end-use environment.”

“This also requires understanding the impact on the material during processes like AFP [automated fiber placement], stamping and welding,” Denize continues. “We also need to define the optimal process window for each material, process and part. So, we have developed tests and simulation models that enable us to do this quickly and thus reduce the development and manufacturing cost for our customers.”  

Patel notes that during traditional TPC part development, “we do several trials to get the right temperatures and pressures, but it takes time. The knowledge base we have developed in CARAC TP allows us to avoid this because we have the fundamental material understanding and computer models to extend that into new parts or to compare with similar materials.”

“We want to stop just doing cooking,” says Michael Hugon, intellectual property manager for Daher. “This team has generated materials science as well as models, rules and methodologies for design and verification that replace the old recipes with a physics and data basis from which we can define new recipes to enable new applications for composites. Instead of saying we have a single competence, such as stamping or welding, we can now say we have the knowledge to look at all materials and develop the part from beginning to end. We can pick the right material, the right process with the right parameters and know how it will perform in the end-use environment. This saves our customers time and money.”

Current certification criteria do not allow TPC materials to be used near their T_g, and non-quality costs for TPC parts can be very high.

TPC technology leadership

Daher (headquarterd in Orly, outside Paris) is the largest independent Tier 1 supplier of metal and composite aerostructures in France as well as the OEM for TBM and Kodiak single-engine turboprop aircraft. Daher’s thermoset composite parts range from main landing gear doors for the Airbus A350, tail booms/fenestrons for the Airbus Helicopters H160, wing panels for the ATR 42/72 and winglets for the Gulfstream G700 business jet, to high-temperature/ propulsion applications such as APU plenums for Airbus, Boeing and Dassault Falcon business jets, a one-piece barrel and strut panels for the LEAP 1B aeroengine nacelle, inner bypass ducts for Pratt & Whitney 812/814/815 engines and a rear secondary structure for the A320 and A350.

In 2009, Daher acquired Airbus’ TPC facility near Nantes, France, and became what it says is one of the first high-volume producers of TPC clips and cleats for the A350. It also supplies the large A400M cockpit floor. Daher has invested in its TPC technology leadership, demonstrating 2-meter-long, high-load wing ribs for the Airbus Wing of Tomorrow program and a curved air intake bulkhead for the 3.56-meter-diameter Rolls-Royce Ultrafan aeroengine. In 2019, Daher acquired thermoplastic composites welding specialist KVE Composites (The Hague, Netherlands) and launched an R&D project to develop a full-scale induction-welded torsion box sized for the TBM aircraft (see “Thermoplastic composites welding advances …”). In 2022, Daher announced a 3-year project with the Luxembourg Institute of Science and Technology (LIST) to mature and optimize infrared welding as a complementary technology suitable for welding thick parts with large dimensions in high-volume manufacturing (for more details, see  “Thermoplastic composites welding advances …”).

In November 2021, Daher completed construction of the 1,600-square-meter Shap’In TechCenter, next to its Nantes composites factory, with the goal to consolidate and accelerate its leadership in composites aerostructures and advance its ability to put designs for aircraft wings, tails and engines into production faster and with greater agility, reducing cost. (Watch for CW’s tour of Shap’In and the Nantes factory in November 2023). This technical center is one piece of a broader strategy announced again at the 2023 Paris Air Show, which includes the Imagineering by Daher open innovation program, Daher’s technology leadership in TPCs and “Take off 2027” strategy, by quadrupling its R&D investment over 5 years with the goal to decarbonize its business and its customers, reaching net zero by 2050.

CARAC TP has provided a clearer understanding of process materials and which materials best fit which applications. It is enabling predictive models to simplify testing and provide rules for AFP, stamping and welding.

Objectives and approach

“We started this project 5 years ago working inside Daher and making tests with French universities and also with the University of Washington [Seattle, Wash., U.S.],” says Patel. These tests, along with various analysis methods, were guided by the following questions/objectives:

• Are TPC materials suited for wings and aeroengine/nacelle parts?
• How do we achieve TPC process control and reduce production cost?
• What are TPC material behaviors above T_g and operational temperature limits?
• How do we build a certification approach?"

The last two are key, explains Hugon, “because current certification criteria don’t allow use of these TPC materials close to their T_g.” Another main focus within CARAC TP was process.

“The goal is to be able to master various processes and reduce the cost of our parts,” says Denize. “For example, non-quality costs for TPC parts can be very high. The data generated by this project have given us a clearer understanding of the process windows for these materials — by which I mean the temperature, pressure and time ranges acceptable within the production processes. It is also helping us identify which materials are the most appropriate for each type of aircraft component: fuselage and wing, engine environment, interior versus exterior components, etc. We have also used the same data to build predictive models that will simplify our testing campaigns and provide rules for our process teams, including AFP, stamping and welding.”

Samples were prepared from high-performance PEEK, PEKK and LMPAEK thermoplastic polymers in unidirectional (UD) tape and fabric as supplied from material manufacturers and tested for mechanical strength and resistance to creep, fatigue, fire, environmental aging and impact — including tests under loads at elevated temperatures. The project also studied the effects of production processes, analyzing material physical chemistry. The results obtained for each material were then compared with those of a reference — AE250 LMPAEK 143 gsm UD tape supplied by Victrex (Clevelys, U.K.) — used because its processability window best suits the production processes originally targeted by Daher in CARAC TP.

Material analysis using flash DSC

The first phase of the project tested TPC tape and fabric prepreg samples using differential scanning calorimetry (DSC) and flash DSC systems. These measure the heat flow to or from a sample as a function of temperature and/or time to quantitatively measure physical transitions and chemical reactions. “Flash DSC allows us to measure process changes at high temperature rates, such as 2000°C/minute, compared to the maximum for normal DSC, which is 80°C/minute,” says Patel. “The TPC materials we tested require high temperatures for melting, but then cooling is crucial because it determines the crystallization which provides the material’s high-strength properties. Flash DSC allows us to understand the boundaries for crystallization — where it occurs and where it does not.”

“We observed the crystallization peak for different temperatures, times and cooling rates,” she continues. “And then we built a time temperature transformation [TTT] diagram with multiple curves. Each curve is a model that enables us to predict the crystallization of the material at a certain temperature and time. So, it defines the process window and helps our manufacturing teams know exactly what to do with each material to provide the properties requested by our customers.”

Hugon explains further, “Imagine we are developing a part and have several processes we could use, including different cooling rates and ways to apply pressure. This type of model covers all processes (from AFP to welding) and allows us to visualize these options and understand their impact on the material’s crystallization. This helps us to choose the most promising process options for the best part performance and also identify the optimal process parameters faster avoiding material degradation. Thus, before even starting work on the real part, these databases and models reduce the time and money required to define the best process and conditions for this part.”

Validation using laminate panels

After completing the initial study on the raw TPC prepreg, Patel explains the second step was to validate this data using tests on laminates processed with different times and temperatures. “We used DSC to analyze samples from these laminates and were able to see the crystallization, but also define the lower temperature boundaries at which we can still achieve the desired crystallization.” This type of analysis can also be done for dwell temperature. “This is important for thick laminates — for example, up to 28 millimeters — which have a thermal gradient inside during processing,” she adds. “We must ensure that if we have such a thermal gradient, crystallization still occurs and is homogeneous along the laminate thickness.”

To measure the temperature inside the parts, the CARAC TP team placed thermocouples at different locations inside the laminates being processed. “So we knew what the temperatures were inside when we applied heat on the outside,” says Hugon. “We also developed simulation to understand the real gradient in between these measuring points. And this is a challenge, depending on the different types of parts. For example, when we are developing large and thick ribs for the next generation of single-aisle aircraft, it’s very important to understand such temperature gradients and discern the real impact of the processes if we want to go to certification.”

High-temperature degradation of materials

The CARAC TP team also analyzed if there is any degradation in the materials after very high temperatures during processing. “For this, we used DSC to analyze laminate samples at several temperatures and times,” explains Patel. The DSC results at right show virgin, unprocessed tape (“Tape vierge”) in green, 480°C in blue, 420°C in red and 380°C in black — processed at 4, 10 and 0.2 minutes represented by solid lines, smaller dashes and larger dashes, respectively.

“We observed that when the material is exposed to different temperatures, there is a shift in the curves versus the virgin material,” she says. “If there is any shift for the same material batch compared to the raw material, then we can assume there is some material degradation. For example, for the longest exposure at 480°C, there is no upswell in the curve that represents crystallization — so there is degradation.”

“We wanted to have a method that enables us to know exactly if we have a degradation or not, depending on the time and temperature,” adds Denize. “At 480°C, we can see there’s definitely a degradation and also some for 420°C, but at 380°C it looks fairly good depending on the time. We want to see the boundaries of where and when this degradation occurs, and maybe to push the limits. This analysis enables us to understand and define the threshold of degradation. We need to understand where the limits are and maybe we need to challenge the limits compared with what we are doing right now.”

Applying HT model to multiple processes

Patel notes this analysis can be used for multiple processes, including AFP, stamping and welding. “For example, when we use AFP to lay up the tape, we sometimes exceed the melt temperature locally in certain areas. To better understand this, we extracted some tape after layup with different temperatures and layup rates and then did DSC analysis but also other chemical analysis — including infrared analysis and chromatography — to identify if there is a chemical bonding change. We can thus generate a diagram with the AFP process parameters compared to the raw material and identify where there is degradation from the chemical analysis.” For example, in one curve the team could see a double peak at a lower level versus the normally tall single peak. “This means that during the processing, we have degradation of the material,” says Patel. “So, this is a key analysis, and it helps our AFP team to control their process window.”

“We can thus generate a diagram with the AFP process parameters compared to the raw material … and it helps our AFP team to control their process window.”

“This is an illustration of the method we developed based on degradation analysis,” says Hugon, “and how we can put it with the process data. We can do this with welding, stamping or whatever process we are using. The goal is to avoid empirical trials, which are very expensive and time-consuming. Having this understanding at the material level with these physical models enables us to choose the right parameters and get good parts very quickly. We will also work with this data in our program with LIST to develop infrared welding, because the target in that project is welding thick parts. So, understanding where the temperature and time limits are and their impact will be crucial.”

HT model for in-use temperature

One of the original objectives for CARAC TP was to answer if these advanced TPC materials can be used above their T_g and to define limits for their use near engines. The team used AFP and autoclave consolidation to make panels and then exposed samples to temperatures between 150°C and 300°C for different time periods. “We then applied the thermal and mechanical analyses we have developed and used those results to create a mathematical model,” says Patel. “We can see, for different times and temperatures, when there is degradation of the material and for different thicknesses. Depending on the criteria we take into account for each application case (mass loss, loss of rigidity modulus, concentration of degradation by-products, thermal properties modification), we have models that can determine the use period for the material at temperature above T_g.  For degradation inside the material, we established a threshold of 5% decrease in modulus, for instance. If we see less than that, we can use the material at temperatures higher than its T_g.” Using this kind of threshold, CARAC TP found for the PEEK, PEKK and LMPAEK materials, time limits for exposure to temperatures above 150°C.

“For each in-use temperature, we can give a maximum time that the material can be used before degradation occurs,” says Patel. “For a 4-millimter-thick quasi-isotropic panel, we saw that we can use the material at 180°C for 100,000 hours with no change in it, but if we go to 230°C, we can use it for 12,700 hours without any change.”