European Commission logo
English English
CORDIS - EU research results

Selective Tape-Laying for Cost-Effective Manufacturing of Optimised Multi-Material Components

Final Report Summary - STELLAR (Selective Tape-Laying for Cost-Effective Manufacturing of Optimised Multi-Material Components)

Executive Summary:
The aim of the Stellar project was to develop the manufacturing processes for high-speed placement of fibre reinforced matrices, in selected locations in a composite structure, to provide the optimum reinforcement, weight and cost profile with a part.
The ideal composite structure has different combinations of composite materials in predetermined locations in a hybrid multi-material structure, but to date this has not been achievable because cost-effective manufacturing processes have not been available.
The concept of this project was therefore to develop the design methodologies, manufacturing processes, equipment and control systems needed for localised placement of different fibre-reinforced thermoplastic composite tapes onto different substrates, creating locally reinforced components that were fully weight-optimised.
To achieve this, the project focused on development of the Automated Tape Laying (ATL) process to selectively place reinforced thermoplastic tapes in 3 manufacturing routes:
• Selective reinforcement of existing components
• Direct additive manufacture of components
• Manufacture of selectively reinforced tailored blanks for compression moulding

Project Context and Objectives:
The use of composite materials in structural components is becoming well established in a range of applications, and the materials can offer significant benefits in mechanical properties, weight and through-life environmental impact. However, whilst some hybrid structures do exist, these are typically combinations of relatively isotropic materials, which only go part of the way to optimising the amount of materials used in the structure.
The manufacturing process developed within the Stellar project will have a significant effect on the weight of structures, as for the first time it will allow different reinforcement fibres (polymer, glass, carbon) to be used synergistically in the same thermoplastic polymer matrix, to produce hybrid multi-material structural components with optimised performance and weight.

The overall goal of this project is to develop the tape-laying process to cost-effectively place different reinforcements locally within a single component, minimising component weight and maximising performance. To achieve this, activities are needed in materials development, process development and machine configuration. 8 primary objectives have been established:
1. Ensure the industrial applicability of the selective tape-laying process
2. Development of Novel precursor materials for the selective tape-laying process
3. Maximising the speed of the tape-laying process for combinations of different reinforcements
4. Downstream post-forming processes to increase component complexity and production rate
5. Synchronous robotic placement of material in complex multi-material structures
6. Demonstration and validation of the selective tape-laying process
7. Assessment of economic and ecological improvements
8. Raising awareness of the potential of the process to reduce material use

Project Results:
The specifications of the materials and process parameters that the project would focus on were defined in the early stages of the project. This ensured that the project outputs would meet existing and planned standards.

Our activities led to the successful production of fibre reinforced thermoplastic tapes. These tapes can be produced in different width up to 30 mm depending on the application. The tape can be further processed by ATL (automated tape laying) into organosheets or complex shapes. It can also be used for localised reinforcement of structural parts. A market study of aligned fibre reinforced thermoplastic tapes was performed in order to continue commercial activity with the Stellar consortium after the project.

The technology and process enhancement of the automated tape placement in particular of the laser-assisted tape placement for enabling a fast, reliable and sustainable production of thermoplastic tailored blanks and multi-material thermoplastic composite components is the overall key result of the work related to the work package concerning the “Direct Manufacture of Components and Preforms”. The implementation and accomplishment of this overall result is outcome of multiple well-defined main results reached by successful conduction of six research and development tasks.
The first main result derived from the initial development task of the work package is the knowledge and feasibility to produce 2dimensional, 2.5dimensional as well as 3dimensional multi-material components. These parts are entirely built out of unidirectional continuous fibre-reinforced thermoplastic tapes in a single process step using laser-assisted in-situ tape placement. Based on various process investigations for evaluation and determination of the most suitable process parameters for combinations out of glass fibre reinforced tape and carbon fibre reinforced tape could be determined. Additionally a systematic parameter variation approach and a well-suited evaluation procedure could be established. In this way process investigations to combine various tape materials using laser-assisted tape placement can be carried out in a more efficient way. A 2d laminate out of glass fibre-reinforced tape combined with carbon fibre-reinforced tape with different fibre orientations manufactured and fully consolidated by using laser-assisted tape placement demonstrates the capabilities. Further possibilities to manufacture also 2,5d and 3d components in a single process step can be derived from this output.
Towards a more efficient process the proof for the possibility to place and consolidate tapes with high process speeds of up to 800 mm/s is the main result of the following R&D task. Tailored blanks are produced using high process speeds for subsequent thermoforming operation. For achieving sufficient consolidation quality considering the follow-up thermoforming process the closed-loop temperature control was improved. In this way the laser system responses faster to the thermal process conditions even at higher speeds. By at least partly consolidating the tape to the substrate material and evaluating the bonding properties after a thermoforming operation, the result could be shown that high speed tape placement in combination with thermoforming delivers as good bonding properties as the slower process for achievement of full in-situ consolidation. Demonstrated by a produced tailored blank with multiple fibre orientations fully consolidated to a random fibre reinforced basic substrate material the feasibility to combine materials with completely different thermo-physical and optical properties using tape placement with process speeds which are higher than the known state of the art processes could also be proven. From these results the potential for high-volume production can be concluded.
The main result delivered by the fulfilment of a development task which focused on the “Selectively reinforced polymeric parts” is a multi-layered result which enables the local application and consolidation of tapes onto a thermoplastic composite part, so called substrate material, at predetermined areas without the generation of waste or the need for subsequent trimming operation. By in-situ consolidation of small or long tape pieces onto a thermoplastic part using selective laser-assisted tape placement local areas of the part can be mechanically enhanced. Thus, the overall part can be designed to the average load case and the load bearing areas can be locally reinforced with tape. This approach enables to save weight, costs and production times. Drawback prior to the carried out development was that the tape tips of the locally applied tape could not be consolidated to the final part by using selective laser-assisted tape placement. These needed to be trimmed away manually after application of the tape. Findings of another task eliminated this additional treatment step. The multi-layered result which enabled us to reach these developments contain systematic concept developments towards a waste reduced additive process, system and control developments for the realisation of the cut-on-the-fly process and finally deep process investigations for a understanding in the effects of local reinforcement onto random fibre reinforced substrate materials. Based on an infrared tape-placement set-up, which enabled a principle development under safe conditions for the worker, concepts for feeding the tape to the processes without generating additional waste and with directly attaching the tape to the substrate were developed, tested and evaluated. Furthermore an active cutting unit based on the principle of a guillotine knife has been developed and implemented to the laser-assisted tape placement head of Fraunhofer IPT. This cutting unit enables a fast cut of the tape during lay-up. Combined with active trigger signals a cut-on-the-fly process which consolidates the tape end to the substrate material could be realised.
Both these developments were key-results towards a waste-reduced selective tape placement process for local reinforcement of thermoplastic composite parts. After implementation of the findings, process investigations which focused on the combination and processing of different tape widths and the influence of using different layup patterns were done. These investigations focused on the local reinforcing a random glass-fibre reinforced black coloured thermoplastic substrate (GMT – glass mat thermoplastic) with carbon fibre-reinforced tape. In this way, results and effects on the mechanical performance when changing the tape width or using various layup strategies could be derived. As results, the foundation towards the automotive and generic demonstrators could be established.
The work in the development under the topic “Catalogue of multi-material interfaces” focused on the interaction of multi-material interfaces when considering the combination of different tapes or when combining tape with thermoplastic organo sheet materials. The results elaborated delivered a better understanding about bonding mechanisms reached by continuous welding. Furthermore, secondary effects influencing the tape deformation behaviour during placement have been identified. The knowledge gained in detail is:
• Beside typical discussed effects of intimate contact and polymer healing, melt flow conditions in the contact region are of high importance.
• The combination of rigid fibres and low viscosity melt supports melt mixing effects.
• Local heating is highly affected by the type of reinforcement.
• Due to enforced melt mixing in the contact region a good mechanical interaction of typically non-miscible polymers can be reached.
• Well adapted process parameters are required.
• In case of material hybrid bonding similar viscosity ranges are required.
Based on these findings material combinations which are highly beneficial for the final product underpin the flexibility and the potential for selective tape placement. Especially when combining the results of the multi-material interface catalogue with the results towards the high speed tape placement and the selective reinforcement of polymeric parts industrial highly relevant manufacturing processes for high volume production as well as for small serial production of high performance products can be realized.

Based on existing simulation tools of the ESI group the results of the development task with the name “Numerical modelling of the ATL process” were initial findings for the numerical simulation of the laser-assisted tape placement process. Key knowledge and existing solutions for processing thermosets were reconsidered and modified with the aim to model the thermal behaviour during laser-assisted tape placement, but also the mechanical behaviour of the applied tape during the application.
Testing and evaluation of the thermoplastic materials has been a key-action of Stellar. Testing and performance analysis should ensure that the selective tape placement process can be evaluated. As the influence of process parameters on the bonding properties when producing multi-materials is important for setting up the process, suitable test methods are needed. Only reliable testing and evaluation methodologies allow a deep analysis of the performance of the part as well as of the functionality of the process. An obstacle for a comprehensive use of thermoplastic composites in particular of thermoplastic tape in various industrial segments is the lack of existing standards and tests to evaluate the materials as well as the processes. Currently used measures are mainly based on testing methods which were adapted from thermoset composite testing. In this context the final development task of the described work package evaluated different testing methods for the characterisation of the Stellar processes and materials. Key result is that various tests deliver results to evaluate single test series. Especially peel tests (wedge peel test, 180°-peel test, mandrel peel tests) can be used for determination of suitable process parameters within a single test series as the results are only comparable if they were elaborated under “similar” conditions. For elaboration of mechanical performance of the materials only the tensile testing according to DIN EN ISO 527 and the 3-or 4-point bending tests following DIN EN ISO 14125 can deliver reliable and comparable results. Other tests like the 3-point bending tests on a short beam following DIN EN ISO 14310 can only deliver indications about the part properties, but cannot be quantitatively analysed as foreseen in the standard. Due to the ductile behaviour of the thermoplastic matrix material the key conditions for this standard which are based on the brittle characteristics of a thermoset-based composite do not apply. Thus, the type of fracture and the measured curve cannot be evaluated according to the standards. Overall suitable testing methods to evaluate the used ALT processes could be set-up, evaluated and used for the project. Based on these finding new test-set ups were improved during the project duration.
The development of component manufacturing processes, by stamp forming and compression over-moulding for enabling a fast, reliable and transferable process chain for processing tailored blanks and sheets was a key aim within the demonstrator production of STELLAR.
Tailored blank preforms with local reinforcement in various fibre directions were taken to process them into a finished component, using either stamp forming or compression over-moulding. Appropriate to the part manufacturing practical and numerical models were developed to simulate part behaviour and process technologies.
Main goals for moulding and forming were:
Post Processing with Conventional Moulding Techniques:
• Determination of the moulding parameters for tailored blanks and sheets for use in stamp forming, using a component representative of key structural features, with a target cycle time of less than 3 minutes.
• Development of compression over-moulding, with a target cycle time less than 3 minutes, to introduce features such as ribs and bosses that are not possible in the stamp forming process.
• Modelling of the complex moulding processes to understand parameters for later demonstrator component manufacture.

Stamp forming of consolidated sheets and tailored blanks:
A task was to produce stamp formed shapes using tailored blanks out of Direct Manufacture of Components and Preforms, to give shaped shell structures with localised reinforcement. The work was based on the prototype demonstrator “Calotte” with a geometric solid corresponds to a sphere geometry consisting of a semi-crystalline polymer (polyamide 6) with glass and carbon fibres used for reinforcement. With that part the drapery and the movement of the UD-Tapes during the forming process was determined which could be transferred to simulation trials. Therefore a special material layup for manufacturing the tailored blanks for the Calotte was defined. Several UD-Tape layers in 0° and 90° fibre directions were bonded and consolidated onto the substrate. The substrate plate for the ATL (Automated Tape Laying)-process was made out of chopped glass fibre UD-Tapes. This plate reflects the upcycling of waste material out of the UD-Tape production. The substrate is used as an underlay for the UD-Tape-Layup and to realize later local reinforcement. The purpose of the Calotte was to determinate the processability and process errors of the UD-Tapes.
The sampling of the Calotte was carried out in two routes. As part of the first route the UD-Tapes-Layers are on a silicone punch of the upper tool. Within the second process route the UD-Tape-Layers are on the polished steel side of the mould. Forming with rubber moulds has proven to be highly advantageous, as it assures an even distribution of pressure within the cavity and the possibility to work with different thicknesses of the sheets/blanks.
For heating up the semi-finished materials a two field infrared drawer oven was taken at HBW. Due to the vertical and parallel arranged heating fields the materials can be heated very homogenous. Between the two heating fields a sledge is mounted on the oven which enables the retractable movement the material. In order to determine important process parameters like heating time and temperature, handling time and cooling time of the materials, heating tests were performed. Therefore material test pieces were prepared. While heating the material, sensors measured the temperature on the surfaces and in the middle of the material. Once the maximum temperature was reached in the core of the material the heated test piece was removed out of the oven and was placed on an aluminium plate for cooling. From this point the handling-time begins and it ends when the melting temperature of polyamide 6 is reached. This time frame represents the time how long the semi-finished product can be transferred in room temperature before reaching solidification.
Manufacturing the stamp formed shapes in the Match Moulding Process a process chain was defined. The fundamental stages are heating the composite sheet, transferring and positioning in the tool, forming the part and cooling the mould, then removing and if needed finishing the product.
Due to transferring the heated sheets from the oven to the tool by hand there was a challenge to place the semi-finished exactly and reproducible onto the mould. Handling systems for composite materials as needle grippers and vacuum suction cups can bring significant improvements to the transferring process. Results of various handling tests show that the material can easily picked up by both handling systems.
On the basis of these pre-trials it was possible to define a whole stamp forming process to produce demonstrator parts with a special steel/rubber tool. Altogether a number of Calotte parts was produced to identify the challenges like positioning and handling time within the manufacturing process and to examine typical error patterns of the parts. Material errors during the process constantly appeared on convex and concave side.
On the convex side the UD-Tapes overlap longitudinal to the fibre direction, this causes wrinkles because the overlaps gets crimped. Transversal to the fibre direction the UD-Tapes are clinched. The result is that the clinched tapes overlap to wrinkles and gets crimped. Other errors on the convex side are distortions transversal and longitudinal to the fibre direction and poor consolidation of small areas in the UD-Tape Layer. On the concave side similar errors occur. Wrinkles longitudinal and transversal to the fibre direction arise because of the same reasons as on the convex side. Furthermore transversal to the fibre direction the UD-Tapes shear away from each other, so that the underlying UD-Layer appears on the surface of the part. Because the UD-Tapes slip out of place transversal to the fibre direction distortions on the surface are noticeable.

Compression moulding over tailored blanks to create complex shapes:
This task concentrated on the compression over-moulding of the tailored blanks to produce complex components with functional elements such as ribs, clips and metal inserts. The work is based on a prototype demonstrator “DropTestPart” (DTP) with a geometric solid corresponds to a U-Beam geometry consisting of a semi-crystalline polymer (polyamide 6) with glass and carbon fibres used for reinforcement. With this part HBW gained knowledge the draping behaviour and the movement of the UD-Tapes during the forming process which could be transferred to simulation trials. Parameters of the compression moulding process varied and assessed, to explore the effect of sheet temperature, flow compound temperature, processing time and pressure, and the adhesion and performance of the over-moulded FlowCompound. For the DTP a special material layup for manufacturing the tailored blanks was defined consisting of several UD-Tape layers in 0°, 90° and ±45° fibre directions which are bonded and consolidated onto the substrate. The substrate plate for the ATL (Automated Tape Laying)-process was made out of chopped carbon fibre UD-Tapes. This plate reflects the upcycling of waste material out of the UD-Tape production. Just as in the previous task, the substrate is used as an underlay for the UD-Tape-Layup and to realize local reinforcements later. The purpose Drop-test part was to determinate the processability and process errors of the UD-Tapes and the processing of FlowCompound materials to realize stiffening ribs or functional elements.
The forming tool which is used for the over-moulding trials is composed of an upper tool and a lower tool. The special feature of the tool is that the die has two lengthways indentations for forming stiffening ribs. Both tool halves are made out of aluminium.

Again, same methods as in the previous task are used in order to determine important process parameters like heating time and temperature, handling time and cooling time of the materials. Furthermore the gained knowledge out of the Calotte part manufacturing was had influence on the processing strategies.
Manufacturing the compression over-moulded parts with the TFC Process a process chain was defined. The fundamental stages are heating the composite sheet and the Flow Compound, transferring and positioning in the tool, forming the part and cooling the mould, then removing and if needed finishing the product. An improvement in reproducibility can also be provided here as in the previous task by the use of handling systems as needle grippers and vacuum suction cups.
Result of this task is that it was possible to define a whole TFC process to produce demonstrator parts with a special aluminium tool. Altogether a number of DTP were produced to identify the challenges within the manufacturing process and to examine typical error patterns of the parts.
Detachments of the outer UD-Tape Layer could be identified as a mainly occurred error pattern of the parts, which can be caused by two reasons. First reason is that volatile ingredients of the matrix can’t pass the layer, so these particles form bubbles and air locks under the tape layer. The other reason for the detachments could be the negative coefficient of thermal expansion, which leads to a shrinkage when the part is heated. While cooling the part the carbon fibres increase to their originally length and detach from the layer below.
These tasks were performed at HBW-Gubesch Thermoforming GmbH. HBW-Gubesch Thermoforming GmbH is engaged in the production of ultra-lightweight, tough structures in high-performance fibre composite and sandwich parts for use in sports articles and automotive. Furthermore HBW is developing lightweight parts for seating and energy absorbing parts out of thermoplastic materials with local reinforcement for automotive applications.

Numerical Modelling of Post Processing:
Within this task practical, numerical models were developed to simulate the structural properties of the DropTestPart and the Calotte and the forming behaviour of the used materials. This allowed us to develop the process technologies to wider applications and industrial processing techniques.
A model for the simulation of compression moulding includes shrinkage models as well as models for deformation resistance. The model describes formability, including surface imperfections and a prediction of geometric distortions.
The main objective of the simulation is to compare the real part manufacturing process with the simulation trials to model the complex moulding processes to understand parameters for later demonstrator parts. Results of the DropTestPart show that no wrinkling is occurring during the forming. However, additional post-treatments as strains in fibres are available. A local maximum value is reached on the radius due to neighbouring plies tapes displacements. A consequence is that manufacturing defects could appear on this zone.
Against that, the simulation of the Calotte part shows major wrinkling because the displacements of the tapes are limited by the sticking contact. Furthermore the total thickness of volumetric elements is simulated. Thickness reduction occurred where forming material presented main wrinkles. Due to the great hardness of the silicone the thickness reductions might reduce drastically according modifications of material parameters.
This task was performed by ESI. ESI is a pioneer and world-leading provider in Virtual Prototyping that takes into account the physics of materials. ESI boasts a unique know-how in Virtual Product Engineering, based on an integrated suite of coherent, industry-oriented applications.

At the start of the project a CAD/CAM simulation environment was configured based on CGTech’s existing software modules. There were VERICUT Composites Programming (VCP) and VERICUT Composites Simulation (VCS). These software modules were developed for Automated Fibre Placement (AFP) and Automated Tape Laying (ATL) based on several different manufacturers of AFP and ATL machines, but had not been configured for use with KUKA robots in a multi robot cell. In addition VCP and VCS had only been used for the layup of complete composites structures which required curing in an autoclave. This process is referred to as Thermoset.
The STELLAR project was to use a head which did not require curing in an autoclave, but was to be cured using a laser mounted on the layup head. This process is referred to as Thermo-plastic. VCP and VCS were developed for layup on complete structures whereas the STELLAR project was formed to investigate local re-enforcement of existing parts or structures with more than one material in a part. So the initial achievement was to build a single robot cell to allow the local re-enforcement of a simple part using the AFPT thermo-plastic AFP head.
This initial virtual cell was then used to develop methodologies for Multi-Material Tape-Laying and investigate the benefits and limitations of variations of head type, robot and control systems.
Another achievement was to develop a methodology to create CAD models of the component prior to the additive process and the definition of the material to be added. This was achieved by CGTech engineers designing a simple test part in a Dassault Systemes CATIA CAD system and adding the geometry and definition of the material to be added.
The initial virtual cell was extended to allow experimentation with two, three and four robots with different heads. However, the results achieved indicated that no real benefit could be achieved by using three or four robots, unless the components were very large and there was sufficient room to allow more than two robots to be active at the same time. The collision detection algorithms on VCS were critical in this phase of the project and the value of having a virtual manufacturing cell was proven.
Having concluded that two robots was the optimum configuration there was further experimentation with different layup strategies including:
• Two robots with the same head operating on different parts of the same component.
• Two robots, each with a different head, laying up material on different parts of the component.
The CGTech VERICUT virtual cell was used extensively by AFPT to finalise the design of the test cell that was to be used in the production of the demonstrator parts.

The VERICUT Virtual cell was used at several key stages of the [project to illustrate key features. VCS allow the capture of both static screenshots and live video which proved invaluable as a dissemination tool.

The objective of demonstration work package was to define and then fabricate case study parts using the information generated in the process development work packages. During the Stellar-project the tape-placement technology was optimised and new tools were developed. It was difficult to demonstrate all improvements in only one or two demonstrators. Therefore, it was agreed to expand the scope of work from two to four demonstrators, but to focus each demonstrator on specific aspects of the project results.
Four different focus markets were defined:
• consumer/sports: production of a demonstrator
• aerospace market: production of a wing part
• automotive market: enhancing random fibre composite
• generic markets: enhancing fused deposition modelling parts

Focus Market: Consumer / Sports
Target: Prove production capabilities for thermoforming with waste optimized tailored laminates
HBW Gubesch introduced the skateboard deck as a concept intended to demonstrate both the thermoforming of waste optimized tailored laminates and the combination with flow compounds.
We learned:
• Two process-routes were compared:
o Local reinforcement of organic sheets before forming
o Local reinforcement after forming
Conclusions: For this product local reinforcement before forming is most promising. This process seems to be highly effective. By using the tape-placement technology with high accuracy tape-placement 30-50 % waste of tapes could be reduced
• Process-information generated by ESI could be used to improve the process
• For skate-board production an effective tool could be developed and produced
• AFPT succeeded to reinforce the organic sheets with UP-tape with limited over-length and without thermal stress
• HBW successfully could thermo-form the tailored blanks
• The focus in terms of the performance of the skateboard demonstrator was to reach comparable mechanical properties like bending and torsion stiffness as a commercially available PP Skateboard Deck (Reference). Therefor HBW performed several mechanical tests with a commercially available PP Skateboard to define the main mechanical properties. The mechanical properties of the demonstrator proved to be close to the commercially available PP Skateboard Deck.

Focus Market: Aerospace
Target: Demonstrate multiple robots communicating to layup on complex surfaces
AFPT proposed a 3D-wing profile as demonstrator case for the aerospace market.
We learned:
• By optimizing its actual tape-placement equipment including the involved software, AFPT was able to achieve full consolidation over the full tape length and to minimize waste of tapes. Demonstrators were produced and showed during the JEC.
• A simple multi-material 3D-form could be produced, ready to produce multiple samples by simple exchange of sheet material.
• The complex 3D-robot-programming only was possible by using CGTech’s Vericut Composites Programming (VCP). This software-tool is able to give support for large parts and multiple robot operation
• AFPT did produce a final demonstrator on a 3d-curved wing-profile, demonstrating accurate tape-placement (tolerances < 1mm.) and eliminating waste of non-consolidated tape-lengths.
• AFPT did erect a multi-robot tape-placement cell with two robot able to work with different tape-heads on one forms.
• For multi-robot-operation the use of CGTech’s VCS-simulation-software is requested to simulate the multi-robot operation and to prevent any collisions between robots and between robot and mould. Although tested on a small scale, no objections are in place to scale-up the process to large forms (e.g. wings).

Focus Market: Automotive
Target: To collect fundamental data to contribute to inner panel for openings and /or closures by local reinforcement
It is in Toyota interest to enhance applications through use of local reinforcements. Before Stellar, the technology readiness level was not suitable for demonstration on large scale complex components. Weak points were tape-placement accuracy and tip-to-tip consolidation. This demonstrator will show the current capability of the technology.
Samples were produced on fibre sprayed material provided by Toyota and 225 tests were carried out.

We learned:
• Sufficient process parameters could be determined for in-situ-consolidation on the fibre sprayed sample material
• Based on these findings a process could be established. Process time was doubles between the first and the last layer.
• Waste of non-bonded tapes could be reduced to a minimum
• Placement process accuracy is good. At the end – after cutting – micro-crimping occurs and cannot be avoided.
• Almost no tape waste is generated during the process.
• Further improvements and mechanical changes could also consolidate shorter pieces of tape to a component and the need for a reconsolidation of the tape can be eliminated with minor system adjustments.

Focus Market: Generic
Target: To demonstrate the potential of Stellar technology to enhance additive manufactured parts
The Stellar project has also been assigned to an additive manufacturing project cluster (FoFAM). In order to enhance dissemination within that cluster, it was therefore agreed within the consortium to introduce a demonstrator that emphasizes the additive manufacturing credentials of Stellar. The demonstrator consists of a substrate fabricated using a fused deposition modelling (FDM) process. The size of the FDM structure is roughly 200mm X 20mm X 15 mm. The demonstrator is locally reinforced with UP-tape.
We learned:
• Tape-placement on the FDM structure could be conducted at a process speed of 250 mm/s without the generation of waste material
• the proof of concept for an economic efficient process for reinforcement of 3d printed components could be delivered.
IPT assessed that:
• the thermoforming process is compared to the key-findings of the thermoforming simulation:
o From the results of the prototyping can be concluded that the processing strategies to automotive, aerospace and other applications are transferable.
o Cycle times can be greatly reduced through the use of automated systems.
o The time for forming, pressing and cooling the part in the mould can be reduced through the use of more effective tempering systems.
o The comparison shows that the forming simulation doesn’t correspond to the real part manufacturing and has to be validated in further work out of the STELLAR project.
• The robustness of substrate materials for the automotive demonstrator was evaluated:
o it is strongly evident that the initial random fibre materials have substantial variance in mechanical properties, but the unidirectional material is far more repeatable in its fibre distribution and orientation than the random fibre material;
o Different simulation environments were tested. Simulations should be considered to be qualitative indicators rather than quantitative results.
o It is expected that further progress in this field can support the development of random fibre material systems with resin and fibre distributions that are inherently robust for the selective placement process or selective placement systems.
• The influences on the material stability during demonstrator manufacturing:
o 4 different tapes are assessed and were analysed against:
▪ the robustness against changing the polymer
▪ the robustness against changing the tape width
▪ the robustness against changing the type of fibre and adding susceptor
▪ the robustness against changing the quality, supplier, fibre distribution, and fibre homogeneity
o we learned that the process of selective tape placement is highly dependent of the tape which is processed

The ecological and economic assessment of the entire lifecycle of a component requires a well-defined methodology following a structured approach. This methodology was developed in detail in this project. The general approach is inspired by the current ISO 14040/44 standard. The methodology was modified by integrating the economic perspective (Life Cycle Costing). Required metrics for both ecological and economic indicators were established considering different attributes like utilizations or cycle times. These metrics were integrated into the Software tool for LCA/LCC (Sophisticated Environmental and Economic Analysis (SEEA) and Quick Scan Methodology (QSM)). The software tool was implemented in Qt/C++. Afterwards, the tool was delivered to TME for testing and evaluation. Fraunhofer IPT received extensive feedback from TME concerning GUI, material database and general usage. It was agreed to adapt the tool according to several points of the feedback. While fundamental changes within the general tool architecture (e.g. further modular layer) were not possible any more, the tool was adapted to be open and now exhibits an interface to read in external material data. Quick Scan Models for Stellar processes were developed in parallel and afterwards included in the tool. In this way, efficient evaluations of Stellar processes are possible as relevant processes are already characterized regarding required inputs like energy, materials or lubricants. Key features of the Stellar LCA/LCC software tool are intuitive (graphical) modelling of process chains, extensive database providing eco-data for all relevant inputs regarded in the Stellar project (energetic inputs, auxiliary and operating materials, raw materials etc.), several eco-datasets (Primary Energy Demand, Global Warming Potential, Ozone Depletion Potential, Eutrophication Potential, Photochemical Ozone Creation Potential and Acidification Potential), provision of process and process chain related eco-KPIs, Hot-Spot analysis, creation of user-specific libraries as well as scenario modelling for automotive and aerospace Use-Phases. For cost evaluations, business scenarios can be defined in order to evaluate process and product specific costs according to customer demands, factory-specific cost rates (overheads, personnel etc.) and shift-models (factory operating times).
Research on recycling methods and applications was performed in this Work Package. This includes general overviews on waste hierarchies and recycling options, e.g. primary or secondary recycling techniques. Furthermore, different industrial applications such as automotive, aerospace, wind energy or naval applications were analysed and discussed considering legislative boundary conditions, common recycling technologies applied as well as current issues and developments. In addition, emerging technologies like eco-composites, self-reinforced composites or eco-design principles were regarded and detailed elaborated. Besides recycling of composites, investigations were done on methodologies for composite repair in this Work Package. In a first step, basic research focused on general characterization of failure and damage in composites and a qualitative suitability assessment on the use of Stellar technologies (selective tape placement) as a repair technique. Besides the evaluation of Stellar technologies for repair purposes other techniques such as laminated scarf repair and patch repair were investigated in detail. This includes general process steps to be executed (preparations, removal of damaged materials, repair inspection etc.) as well as geometric KPIs defining boundaries and requirements for certain methodologies (e.g. “scarf ratio” for laminated scarf repairs). The potential of tape placement for repairing purposes was elaborated and advantages as well as limitations were carefully outlined. In order to execute specific tests and to evaluate this potential on an empirical basis, a dropped weight test up was designed for introducing defined and reproducible damages on components to be repaired in later project stages.
Commercial LCA software and material datasets were purchased for modelling and assessing the ecological impact of Stellar materials for SEEA/QSM software integration (static database) and the general assessment of the Life Cycle Impact. As the processes used in the commercial LCA models were based on averaged datasets sensitivity analyses were performed for determining potential errors. It could be shown that the major impact of composite materials originates from initial granule materials and fibres (particularly carbon fibres). Innovative processes like laser assisted tape-laying / placement were characterized according to the energy- and resource consumption measurements performed at AFPT in an industrial scale in Dörth, Germany. Ecological “impact-drivers” were identified and the results were transformed in models to be used in the Software tool.
On the basis of the process and material information collected during the first six months, an early economic assessment was performed. The Stellar partners delivered cost information for the specific Stellar materials like granules, fibres and semi-finished materials such as UD-tapes. Furthermore, material suppliers were contacted for obtaining cost information which could not be collected within the consortium. On this basis, a first material and process cost overview was established for supporting the development programme and establishing a basis for final evaluations. Furthermore, cost models originating from the tool development were used within this period for challenging the Stellar processes and potential applications with current practices respectively specific cost requirements.
For environmental assessments, further power measurements were performed in a lab-scale in order to identify process-dependent and process-independent consumers and shares (e.g. laser, control system, laser cooling, robot axis etc.). This was not possible during initial measurements in an industrial scale at AFPT in Dörth.
It could be shown, that a major share of power consumption is related to fix consumers such as laser and optic cooling, control systems and basic loads of robots and laser. In the specific case (PA6-CF UD tape, laydown rate: 200 mm/s) process-independent power demand had a contribution of almost 60 % to overall power demands for one sample (l = 250 mm).
On this basis, strategies for increasing ecological efficiency were deducted. The strategy of Highspeed-Placement (increased laydown rates and thereby reduced cycle times) was investigated in detail. For this purpose, an adaptive process chain (tape placement and forming) was designed in order to allow variations of laydown rates. Ecological improvements could be quantified by corresponding eco-KPIs such as Global Warming Potential or Primary Energy Demand.
It could be shown that the process-related Global Warming Potential can be reduced by more than 60 % when increasing the laydown rate from 200 mm/s up to 800 mm/s (for a simplified flat sheet setting).
The automotive Demonstrator was evaluated in detail regarding ecological implications. The increase of ecological impacts due to additional tape layers (selective reinforcement) was compared to corresponding improvements of mechanical properties and current practices (e.g. other materials like C-SMC). In this way the specific “impact per improvement” could be quantified for different eco-indicators defined in the beginning of the project.
An exemplary conclusion of those findings is that “30 Joule additional process energy are required for increasing the flexural modulus by 1 MPa for a random fibre sheet” (PA6-CF, 30 mm fibre length). Those conclusions on “impact per improvement” were evaluated to different substrates (different fibre lengths), different mechanical characteristics (Bending Strength and Flexural Modulus), different KPIs (all 6 eco-indicators defined in Task 7.1) and different placement strategies (parallel to fibre orientation, perpendicular to fibre orientation etc.).
Furthermore, “what-if-scenarios” for evaluating the ecological improvements of lightweight design in the Use-Phase of a car were calculated in order to examine long-term improvements of lightweight technologies as developed in Stellar. It could be shown that consequent lightweight design provides tremendous impact savings within the use phase. As an example, 10 % weight savings of compact car chassis may lead to savings of more than 300 kg CO2-Eqv. (diesel car, reference weight of chassis: 450 kg). This impact is equivalent to the GWP impact included in 6 kg pure carbon fibre. Scenarios were made for different fuel systems and component designs (e.g. adapted gear ratio due to light weight which leads to even more savings per distance).
The same approach was applied for aerospace applications considering short and long distance aircrafts. For those scenarios annual distances between 2 and 6 million km were assumed.
Those investigations conducted for the ecological perspective were also applied for economic indicators in this Work Package. Therefore specific measurements were linked to cost data for different inputs such as tapes, sheets or power consumption. Market research and discussions with material and system provider were executed in order to gain information on purchase prices. Equal Use-Cases to the environmental investigations were regarded.
Regarding High-speed Placement, it could be shown that the process-related costs can be reduced by more than 70 % when increasing the laydown rate from 200 mm/s up to 800 mm/s (for a simplified flat sheet setting).
In addition, different business scenarios for economic attributes like factory operating time, customer demands or in-house tape manufacturing were considered and analysed.

The main dissemination channel for the project has been the website ( and through attendance at events such as JEC World and Composites Innovation. A range of marketing collateral was produced including flyers, posters, postcards, booklets, case study datasheets, infographics and videos. These were utilised at events, conferences and workshops to aid in the effective dissemination of the project.

Potential Impact:
These actions have led to the identification of relevant precursor materials and manufacturing processes for required performance and cost effective manufacturing.
The fibre-reinforced thermoplastic tapes offers a large range of materials for industries such as automotive, aerospace and wind energy in terms of process speed and versatility, cost effective processing and mechanical properties.
As part of the results with high economic and social impact the process of tape placement process has been enhanced to enable the placement of long as well as small pieces of tape without the generation of waste. Automated tape placement process prior to Stellar could not reach a fully consolidated bonding between the applied tape and the substrate over the entirely applied tape length. Hence, additional subsequent trimming operations which are often conducted manually were required. This resulted in economic inefficient and environmental unfriendly production processes for unidirectional and/or partly unidirectional reinforced thermoplastic composite parts. The generation of waste material which is trimmed and thrown away causes unused raw materials and costs (e.g. energy consumption and during material and part production) without adding value to the product. Additionally the trimming operation causes an additional process step which increases the complexity of process chains and the production times of the final part. Due to these drawbacks such production processes were mainly not applicable for high-volume production, such as in automotive production. The results of the work performed in the Stellar project delivers the key to overcome these issues. Not only high-volume production is a key-beneficiary of the development inside Stellar, but all composite manufacturing industry especially aerospace who seek an efficient, fast, reliable and sustainable production methods for high-performance thermoplastic composite part will benefit from these results.
Generally the overall understanding in production of hybrid thermoplastic components comprising two different kinds of composite materials, whilst one of these materials is a unidirectional continuous fibre-reinforced thermoplastic tape, enables us to produce parts with the optimum weight, reinforcement and cost profile. The developed process and system technology allows us to implement and use this key-knowledge in an industrial way. Due to this we outcome low-performance and thereby cheaper substrate materials can be locally reinforced on the load bearing areas using selective tape placement. In this way, the part thickness of the substrate material can be decreased to withstand the average load cases, as the peak loads are taken by the locally applied tape layers. Thereby costs, materials and weight of the overall parts can be saved. The numerical modelling of the process for achievement of these impacts provides the required knowledge for setting-up and understands process principles prior to production. In this way the time-to-market for tape placed hybrid parts or locally reinforced hybrid thermoplastic composite parts is reduced, because the required amount for pre-trials and setting-up trials can be drastically reduced. In this manner materials, energy, costs and time will be saved. Summarising the output of this work provides the potential to have a huge impact on the European composites manufacturing industry as these findings are the key-enabler to produce thermoplastic high performance components in an economic efficient and sustainable way. This impact applies to both high-volumes as well as for small-series production.

By adopting the technologies to make thermoplastic composites with localised carbon and glass reinforcement, HBW-Gubesch will be able to offer their customers a wider range of lightweight components to replace the steel parts currently used. This will enable HBW-Gubesch to enter the mass market.
Besides the economic aspect, the knowledge will be expanded in several ways by this work. The enhancements of knowledge include:
• Expansion of knowledge in stamp forming process technologies and process parameters for manufacturing lightweight components with glass or carbon reinforcement out of tailored blanks with ATL laid tapes and random fibre tapes.
• Expansion of knowledge in compression over-moulding process technologies and process parameters for manufacturing lightweight components with glass or carbon reinforcement out of tailored blanks with ATL laid tapes, Random Fibre sheets and FlowCompounds
• Knowledge of draping and forming behaviour of localised carbon and glass reinforcement during forming processes.
• Knowledge of material behaviour between localised reinforcement and random fibre sheets.
Due implementation of this work the development of the moulding parameters and compression over-moulding, as well the modelling of the complex moulding processes for later demonstrator component manufacture could be fulfilled. Furthermore the process can be made suitable for large series production by using upscale process systems. Firstly, several materials can be heated by a plurality of heating stations and stronger heating systems and secondly, handling times can be reduced through the use of automated systems such as turnstile or material handling systems such as needle grippers or vacuum suction cups.
Regarding the implementation of the results a statement can be made that these various technologies present a high potential for establishment of a series-compatible product in industry. There are numerous companies which have interest to put their experience and technical achievements to the implementation of these technologies. Finally, the application of tape laying will occupy an important place in automotive and aerospace sectors through local reinforcement of composite materials by conventional moulding techniques.

With the creation of the Virtual CAD/CAM environment it has been possible to program and simulate the function of the robot cell in a safe risk free manner. A number of different robot layouts and layup strategies have been investigated in the virtual world prior to the creation of the robot cell. The virtual robot cell has been used to compare alternative methodologies both in terms of cycle time and material usage to establish the best methodology for each demonstrator part. The virtual cell has allowed far more extensive experimentation than would be possible in the real world and the collision detection functions have proved invaluable in allowing the robots to work safely in close proximity to each other. Use of the virtual cell has ensured that the layup test have been conducted safely and with a minimum of material wastage. In summary, the output of this work provides the potential to have a huge impact on the European composites manufacturing industry as these findings are acritical part of the methodology to produce thermoplastic high performance components in an economic efficient and sustainable way. By the delivered results these impacts apply for both high-volumes as well as for small-series production.

Within the demonstration work we were able to show the practical use of the research developed within the rest of the project:
• The numerical modelling tools generated by ESI were used for design of components and to determine process-parameters
• It was demonstrated that with Stellar’s technology, products can be produced which are technically competitive to products on the market
• With the software tools developed by CGTech it is possible to scale up production of thermoplastic fibre-reinforced products
• The multi-robot environment created by AFPT demonstrates the scalability of the process.
• A stable tape-quality is vital to get a competitive output and product-quality.
• The output from the process simulation for thermoforming plus overmoulding process does not correspond the output from the actual thermoforming process. This will be improved in the next years.

As part of the wider impacts final ecological and economic evaluations of the Stellar manufacturing technologies and applications were produced. Although the focus of the Stellar project is on the manufacturing stage, also subsequent phases of the life cycle were carefully regarded. The use-phase was modelled according to weight savings and corresponding decreases in fuel consumptions for automotive and aerospace applications. Repairing technologies as well as End-of-life options were discussed and evaluated for different industries and applications.
The final evaluations could verify and quantify the advantageousness of Stellar process technologies and therefore stimulate dissemination and exploitation activities. Stellar achievements towards cost- and eco-efficient manufacturing of optimized lightweight components could stimulate the general application of lightweight design in different industries such as automotive or aerospace. The Use-Cases could already show that significant ecological and cost savings can be expected for those “what-if-scenarios”. Those efficiency increases could strengthen general market acceptance and the fulfilment of legislative requirements towards greener mobility. The same applies for findings concerning repairing and End-of-life methodologies which are still not fully established in the field of composites. Although the Stellar elaborations in those fields focus on the theoretical background and general conclusions/recommendations on the use of these techniques, progress was made as repairing and recycling of components originating from Stellar technologies is covered. As an example, tape placement itself could be used as recycling technology which again stimulates the application of this process as it provides extended “value” in the field of composite lightweight construction.

The website, along with a strong visual identity, was created within Month 1, and has seen a steadily increasing number of visitors. Over the duration of the project (1st Sep 2013 – 31st July 2016) the website has received almost 16,000 page views from nearly 4,000 individual users. The average length of a visit (avg. session duration) was 2 mins 13 seconds and the bounce rate was 68%. Visitors to the site were truly global with the majority from Europe, North and South America and Asia. Specifically, most visitors were from Russia, United Kingdom, Germany, Brazil, France, United States, Spain, Italy, Netherlands and India. In terms of demographics, the majority of visitors were male while 22% were female – this compares favourably with a recent report showing that only 13% of people working in science, technology, engineering and mathematics occupations were women. For age, a typical normal distribution was seen skewed towards the younger ages with the majority of users (36%) being 25-34 years old.

Stellar and project partners NetComposites, HBW, ESI, AFPT, CGTech and Fraunhofer IPT exhibited at JEC World 2016. The three day show was attended by nearly 37,000 composite industry professionals from more than 100 countries and lead to many interesting discussions and enquires across the 7 exhibition stands. The Stellar project generated much interest with its innovative automated fibre placement technologies at Composites Innovation 2016, 22-23rd June 2016, Sheffield with talks from project partners and an exhibition stand including samples and parts made within the project.
The talks will were:
• Recent Advances in the Automated Processing of Thermoplastic Tapes – Tido Peters, Fraunhofer Institute for Production Technology IPT
• Developments in Programming and Simulation for Composite Lay-up Manufacturing – John Reed, CG Tech
• Hybrid Structures Manufactured by Using Thermoplastic Tape Placement for Local Reinforcement Build-up – Ralf Schledjewski, Montanuniversität Leoben

List of Websites: