Community Research and Development Information Service - CORDIS

FP7

CoVaForm Report Summary

Project ID: 606171
Funded under: FP7-SME
Country: Germany

Final Report Summary - COVAFORM (Conservation of valuable materials by a highly efficient forming system)

Executive Summary:
In times of increased international business competition forging companies try to increase their competitiveness by optimization of different factors such as efficient use of resources. For the forging industry the biggest part of production costs are material costs, which sums up to 50 % in case of steel and more than 90 % in case of titanium for the majority of parts.
Therefore one of the easiest possibilities to increase the competitiveness of SMEs is by saving raw material. The advantage becomes even higher for valuable materials such as bainitic steel and titanium. The aim of the project CoVaForm is a holistic approach in increasing the competitiveness of forging companies by developing technologies for a wider use of resource efficient preforming technologies. Therefor four main project results were achieved covering the range from developing cross wedge rolling (CWR) processes to the analysis of the final parts.
The first step within the project was to develop resource efficient process chains including CWR as a preforming operation. This has been done by the help of FEA simulations. The results show the feasibility of using FEA for developing CWR processes. All of the developed process chains enable a significant flash reduction and are therefore suitable for increasing the competitiveness of SMEs.
The next step was to develop a method for an automated design of CWR tools. Although the CWR technology offers many advantages it has not been widely accepted throughout the forging community. This is due to the complexity of the design of CWR tools. The new method allows the user to easily create CWR tools and has been successfully implemented in a software called ROLLCAD.
Because CWR machines are very expensive nowadays another goal of the project was to develop a CWR machine designed according to the needs of SMEs. Within the project time a machine has been developed enabling SMEs to use the CWR technology. The machine is a flat cross wedge rolling machine and enables easy tool change. Within the project time the machine has been successfully tested at OMTAS. The machine has proven the suitability to be used in a forging environment.
Another task to overcome the reservation of the forging community to CWR technology was to develop a CWR inspection method. The developed method is based on thermographic inspections. By this the CWR process can be controlled and in case of failures the user has the possibility to detect this early on during the rolling process.
Lastly several parts have been forged and analyzed. The results of the investigations have shown that there were no failures due to CWR in the newly developed forging sequences.
Project Context and Objectives:
Whether producing engine parts, turbine blades or hip implants: during production of forging parts with a long geometry material is the main component of the production costs, especially if valuable materials like titanium or high alloyed steels are used. Most parts are forged in dies in several steps with a high amount of flash. One way for saving valuable material especially for long forgings is using cross wedge rolling (CWR) as first preform operation. The CWR tools can be designed in round and flat wedge tool configuration. The main advantage of round CWR wedges (see attachment Figure 1, left) is that a high batch production (mass production) is possible. However, the tool technology is difficult and the round wedge tools are very expensive, depending on size and geometry. The use of flat wedge tools (see attachment Figure 1, right) is an economical alternative due to less expansive tools and therefor attractive for SMEs.
CWR machines for industrial processes are almost exclusively expansive machines with round wedge tools. Especially for SMEs this technology brings along several challenges:
- lack of knowledge concerning the design of CWR processes,
- missing knowledge about the boundary conditions of rolling of valuable materials,
- missing alternatives for economic CWR machines for small and medium batch sizes, and
- no quality control to ensure a fault-free production.
OBJECTIVES
The challenges mentioned above have been overcome by developing a new resource efficient process chain for forging parts made of steel and titanium. The investigated parts were long parts with irregular mass distribution along the main axis (see figure 2). Particularly for that kind of parts the design of a flash reduced forging sequence is very difficult.
In order to achieve significant flash reduction, the process chain that has been developed includes a CWR preforming operation and an adaption of the final forming dies (see attachment figure 3). Thus, a material utilization of nearly 100 % in preforming and up to 80 % in final forming has been obtained.
Designing CWR processes requires a lot of specific knowledge. In order to enable SMEs to use this technology, an easy to use CWR process design method was developed that enables inexperienced users to design CWR processes on their own.
The investigation of CWR for new materials is the second challenge. Since CWR is unknown e. g. for titanium until today, another project objective was to investigate CWR for new materials and to develop a working process by practical tests and the application of modern numerical simulation tools (FEA). The intention was to find boundary conditions in which the rolling of valuable materials is possible.
To overcome the economic disadvantages of expensive machines with round wedge tools, a CWR machine with flat wedge tools has been developed which is economic especially at small and medium batch sizes (see attachment figure 4). The machine was designed accord-ing to the needs of SMEs, such as easy tool change and low tool costs which is this third challenge. The new CWR machine developed in the current project does not need a hydraulic press to maintain pressure. Instead, a hydraulic or electric device will be used to enable the needed press force.
Addressing the importance of stable work piece temperature for CWR a method for reliable measurement was required, which resembles the last challenge. In order to avoid any inter-ference to the CWR process all measurement and inspection methods are supposed to work without any physical contact to the work piece. Thus, thermography was used as process monitoring technology to measure the work piece temperature and, in addition, inspect the CWR parts in order to detect surface defects.
Project Results:
1 Process chain development
1.1 Process chain development for the demonstration part of OMS
OMS forging company proposed several parts for CoVaForm project with the aim of developing a new manufacturing process chain that includes Cross-Wedge Rolling (CWR) technology. Among these parts, the one called "X0512.A" was selected taking into account that it is a part currently being manufactured by means of conventional forging technology. This offers the possibility of making a comparison in terms of material usage between the current process and the one to be developed.
X0512.A (Figure 1 and Figure 2) is a forged part for the agricultural sector that is made of SAE8625H steel grade, usually used in forging pipes, tubes and shafts with flanks. This part has a production rate around 6000 pieces per year at OMS forging company.
X0512.A part is currently forged in two steps: 1) preforming stage + 2) final forging stage. The current forging process starts from a squared bar (side = 110mm; length ≈ 170mm), which is transformed to a complex preform (Figure 2). The geometry of the preform is designed to distribute the material in a suitable way that assures the right filling of dies cavity in the final forging stage and, therefore, that the part dimensions are within the previously defined tolerance ranges. At the same time, it must be also guaranteed that no problem of folds appears due to an incorrect flow of the material in the final forging stage.
The forging process is performed in a mechanical press at T = 1250ºC to benefit from the reduction of the material yield strength when working at high temperatures. However, due to the size of the part (billet mass = 16.28Kg), great forging loads are necessary for obtaining the final shape of the part. In this way, OMS estimates that forging load is around 6000-7000t.
To control material flow and to keep it within the dies cavity, brakes and flash channels are placed all around dies cavity. These elements provoke an increment of the forging load, but, at the same time, they help in reducing the necessary material to fill the cavities (Figure 3). The right design of brakes and flash channels is a key point of the forging process when manufacturing this kind of parts.
Before suggesting an alternative manufacturing process, which includes a CWR operation, it was considered necessary to simulate first the current forging process with the aim of studying the material flow and of identifying the most critical zones of the part in terms of die filling.The process was simulated by means of FORGE NxT 1.0 software, which has been developed to enable the simulation of many cold and hot metal forming processes. Starting from the preform obtained by OMS in the previous forging operation, the material evolution during the final forging stage was monitored (Figure 4).
The material flow shows that the most difficult zone to be filled is the one corresponding to the disk with the bigger diameter and also that flash is quite uniform all around the forged part (Figure 5). Besides, it was checked that the material flows in such a way that no fold appears on the surface or within the part.
As the initial billet mass is 16.28Kg and the final forged part mass is 12.93Kg, around 3.35Kg per piece are transformed into scrap after finishing the manufacturing process. Or in other words, around 20.55% of the initial forging material is finally discarded. Therefore, reducing the flash is the objective to get by means of modifying the manufacturing process. Reducing the initial billet mass impacts not only in the amount of material to be consumed, but also in the necessary energy to heat the forging material from room temperature to the working temperature.
To substitute the current preform by a rolled preform manufactured by cross-wedge rolling technology, it is necessary to define first the number of reductions, the reduction rate and the forming and wedge angles. Taking into account the geometry of X0512.A part, which, for simplicty, could be considered as a long nearly cylindrical zone followed by two disks of bigger diameter, it was decided to look for a preform with just one reduction in the central zone with the objective of obtaining two preforms at the same time, thus increasing the preforms production rate (Figure 6).
To identify the preform able to fill the dies cavity with the minimum mass, while avoiding the appearance of any fold, an optimization study was launched with FORGE NxT 1.0 software optimization tool. This tool allows defining the preform dimensions as the parameters to be modified during the optimization process with the objective of minimizing the preform volume. As it was mentioned before, dies filling and folds must be taken into account in the optimization study in such a way that they are considered as forging process constraints.
Once the optimization results are processed, FORGE NxT 1.0 software shows which are the parameters combinations that fulfill the constraints and, at the same time, it identifies the best combination for minimizing the preform volume. As conclusion of this particular study, the simulation called Gen1Ind2 was identified as the best solution for the cross-wedge rolling preform (Figure 7). Figure 8 shows how the flash distribution is for this preform at the end of the final forging stage.
The geometry of the selected preform is sketched in Figure 9 and the main characteristic is related to the initial and reduced diameters (120mm and 60mm).
Using cross-wedge rolling technology, this preform would be obtained with a reduction rate of 50% and a forming angle Alpha = 13.70º and it owns a mass of 15.91Kg, which means that, when comparing it against the current preform, a reduction of 0.37Kg is obtained (Figure 10). Taking the current flash mass (3.35Kg) into account this reduction is around 11% of its value, which it is a quite relevant reduction, because it would save more than 2000Kg of steel per year (0.37Kg x 6000 pieces =2220Kg).
To check the feasibility of obtaining the proposed preform by cross-wedge rolling technology, its rolling process was simulated and cross-wedge rolling dies were built using a wedge angle (Beta) of 5º (Figure 11). This rolling dies building is a first approach to a future industrial design and they were built with the objective of being used for process simulation purpose.
The objective of cross-wedge rolling simulation process was to confirm that no defects appear in the preform when being rolled. The typical defects, which usually appear when rolling different preforms, are spirals, neckings, grooves, folds and internal cracks (Mannesmann effect) (Figure 12). Any of these defects directly impacts in the quality of the forged part and their appearance means that the affected part is discarded and transformed to scrap.
Due to the dimensions of the part to be rolled (around 600mm of length), the length of the proposed dies are quite important (3425mm), what could be considered as the main drawback for setting-up this manufacturing solution at industrial scale. However, the material flow evolution during the cross-wedge rolling process seems to be correct enough because no external defects appear on the surface of the rolled part (Figure 13). To check the presence of internal cracks would need a deeper study, but, due to the characteristics of the process (steel grade, reduction rate, forming and wedge angles), it is expected that no significant problems are to be found in that sense.
Summarizing the obtained results: although CWR dies would need a more detailed definition, the performed study shows that it is feasible to reduce the material usage in X0512.A part manufacturing process in a relevant way (around 11% of the flash would be saved) if the preform is obtained by cross-wedge rolling technology.
1.2 Process chain development for demonstration part of OMTAS
In addition to the demonstration part of OMS two model products of OMTAS were selected as well in order to develop a new forging sequence containing a cross wedge rolling step as a preform operation. The selected parts are a hip implant made of titanium and a common rail made of bainitic grade steel.
In order to identify the suitable CWR parameters for the different materials, basic investigations were performed to investigate the influence of the process parameters on the rolling result. The tool parameters, which differ from material to material, are the forming angle, the wedge angle and the reduced diameter (see Figure 14).
The process parameters are the forming velocity, the tool temperature and the billet temperature.
First, FE simulations were performed using a wide range of values for the material in order to get the perfect parameter combination for rolling titanium and bainitic grade steel (see Table 1).
The simulation investigated the best parameter combination for rolling valuable materials. The best parameter combination was:
- Forming angle: 30°
- Wedge angle: 6°
- Cross section reduction: 40% - 50%
- High forming speed
With these results, real trials were performed in order to confirm the simulation results.
In total, over 450 CWR trials were performed and investigated. With these tests done, the process layout for the two parts could be performed.

1.2.1 HIP IMPLANT
The conventional hip implant was produced using a two step process chain. The sequence starts with a bending operation in order to get a first simple mass distribution according to the angle of the head of the hip implant (see Figure 16, right) and ends with an open die forging step. The flash quota of this sequence is 69 %. The used billet had a diameter of 28 mm and a length of 160 mm, the sheared weight was about 438 g.
The shown conventional forging sequence was the starting point of the following investigations. As a first step in order to significantly reduce the initial billet material a redesign of the forging sequence was executed.
The general idea of the optimization was to reduce the complexity of the forging part geometry by forging two implants in one forging step. The mass distribution of the final parts is more homogeneously this way and thus the necessary material can be reduced. The two-part design has a mass distribution with a trend line with a gradient of -0.05, for the conventional design the trend line has a gradient of 1.86 (see Figure 17). To improve the material consumption for producing hip implants, this first general redesign was realised and investigated.
To improve the existing forging sequence while meeting the demands of small and medium size enterprises the forging sequence was redesigned using a CWR preforming operation for an improved mass distribution and a die forging operation with flash brakes to enable an improved material flow (see Figure 18).
The geometry of the preforming operation was designed using the mass distribution diagram. Using cross wedge rolling enables a rotation symmetrical preform with a mass distribution adapted to the mass distribution of the final part (see Figure 19). The difference between the CWR preform and the final part is more homogeneous.
With the new final forging die design, the initial billet has a diameter of 30 mm, a length of 129 mm and a weight of about 405 g. The billet after CWR consists of a cylindrical piece with two mass allocations at the ends (see Figure 18). The mass allocations diameter is 30 mm and the reduced diameter in the middle is 20 mm, which is a cross section area reduction of 55 %. The length of the cross wedge rolled preform was 182 mm.
The flash brakes on the final forging die are located so a channel is created to steer the material in the area with the biggest material accumulation (see Figure 20).They are only used partially to prevent the die pressure to increase too much.
FORM FILLING
The cross wedge rolled preform and the final die forging with flash brakes result in a complete form filling (see Figure 20).
The moment when the whole engraving is in contact with the billet material is defined as the moment of complete form filling.
With the newly developed forging sequence, the flash quota could be reduced significantly from 69 % to 32 %. This is a material reduction for one hip implant of 235 g (see Figure 21).

1.2.2 COMMON RAIL
The second part was only investigated using simulations. The conventional forging sequence consists of one preforming step and a final forging step (see Figure 22).
NEW PROCESS CHAIN
For the new process chain, a cross wedge rolling step was included in the conventional forging sequence.
In order to get a suitable CWR preform for the following forging steps, several CWR geometries were investigated (see Figure 23).
With the best CWR preform the process chain was developed and the following forging steps investigated (see Figure 24).
With the newly developed forging sequence, the flash quota could be reduced from 35 % to 30 %. This is a material reduction for one common rail of 1340 g (see Figure 25).

2 Development of CWR machine
The starting point for the design of a cross wedge rolling machine was a systematic picture (see Figure 26 which was presented in the project proposal.
For the horizontal movement, in order to set up the roll gap between the wedge tools, and for the vertical movement, the moving direction of the wedge tools, different drive concepts were investigated and analysed concerning their feasibility for the CWR machine.
HORIZONTAL DRIVE SYSTEM
As possible drive systems for the horizontal movement following possibilities were identified:
- Ball screw drive
- Linear drive
- Hydraulic drive
For each drive system, the pros and cons for the suitability for the CWR machine were investigated and the best solution selected.
For the CWR machine, the ball screw drive was chosen. The ball screw drive system is able to realise the needed forces with a high efficiency (0.95 – 0.99). Additionally the construction of the drive unit is simple and resilient.
VERTICAL DRIVE SYSTEM
As possible drive systems for the vertical movement following possibilities were identified:
- Electric drive
- Hydraulic drive
- Worm gear screw drive
For each drive system, the pros and cons for the suitability for the CWR machine were investigated and the best solution selected.
Even though the hydraulic drive would enable a steady movement with the ability of moving big masses, the worm gear screw drive was chosen. The worm gear screw can realise the needed forces as well but also is more suitable due to the self –locking effect which prevents the upper part of moving downwards by itself. Additionally the complete drive unit is very robust.
FIRST DESIGN
With the chosen drive systems, the first concept was developed (see Figure 27).
The developed concept consists of three layers: the basic ground plate, the moveable middle section with the moveable lower CWR tools and the upper plate with the upper CWR tools. The middle section can adjusted vertically in order to set up the roll gab.
To ensure the stability of the CWR machine during a CWR process, the stiffness was simulated using FEA. The assumed forces were gained in simulations within the CWR process for the two model products were performed. The forces in horizontal and vertical movement according to the simulations were used for the stiffness simulations in order to get realistic results.
Due to these simulations, the machine was adjusted with thicker plates for the machine and with additionally added cross sections to improve the overall stiffness (sees Figure 28).
FINAL DESIGN
Due to the expensive worm gear screw drive, the machine was redesigned significantly (see Figure 29). Since the roll gab needs to be adjusted once for every CWR tool, an automated moving system is not necessary. Therefore the amount of “machine beds” are reduced to two. With this the overall weight was reduced and the total stiffness increased. Instead of moving the middle machine bed, the upper machine bed is moveable in order to set up the roll gap. Also trapezoidal spindles are used for manual movement instead of electrical jacking system. Additionally precision shaft steel with support were used as a substitute for the ball rail system, this enables an easier tool change.
The upper machine bed is moveable using two trapezoidal spindles. In order to get the needed roll gab, for each part different spacers were manufactured to prevent the machine beds in moving to far from each other. These spacers are placed on the upper machine bed and under the guiding column (see Figure 30, red circle).
For this design, for all parts the necessary technical drawings and CAD files were prepared and manufactured. The parts which could be purchased were bought and finally all components mounted together (see Figure 30).
The initial concept had a total weight of approximately 31.1 t, with the different development steps, the total weight was decreased significantly until the final design was developed with a total weight of 12.6 t.

3 Development of Inspection system
In order to increase the acceptance of CWR technology within the forging industry a process monitoring system has been developed. Therefore thermographic inspection has been used. The system uses a camera and processes the thermographic inspections for monitoring the CWR process.
To verify the readings of the camera, the temperature at two points of time are considered: First, the initial temperature right after taking out the piece from furnace and second, after handling the part to the starting position of the CWR machine. The initial temperature is known and is in the range from 1000°C to 1250°C. It takes about 10s to move the part to the starting position. A simple finite difference method (FDM) is utilized to calculate the temperature (see equation 1).
The boundary conditions were set to convective (see equation 2) and radiative (see equation 3) heat transfer at the entire surface.
The calculation is done with thermal conductivity (at 950°C), density (at 950°C), specific heat capacity (at 950°C) and thermal transfer coefficient (steel to air, natural convection) . It is shown by Traxler et al. that the emissivity value of the hot material after a rolling process is close to 1. This assumption is further verified by using relevant experimental methods, namely by taking temperature values from a location, at which an emissivity close to 1 can be expected due to its geometry, as shown in Figure 31.
An emitted ray of infrared radiation (Figure 31) is reflected by the pressing-plate, which acts much like a mirror. When the reflected beam hits the CRW-part again, the fraction that reflects will be increased by the emitted radiation. Due to such multiple reflections, sometimes referred as cavity-effect, the effective emissivity gets close to 1.
Thermal images are captured by camera during CWR-process with a sampling period of 100 ms (Figure 32).
For monitoring the process and controlling the quality of the products, thermal profiles along the axis of the piece are collected and assembled into an image, as it typically is done with line scan cameras.
Figure 33 shows the time evolution of the profile measured along the red line shown in Figure 32 (down). The rectangle in Figure 33 is the area of interest, which contains the temporal window, in which the CWR-process is in progress. This region represents a unique signature that is always observable in normal operation. The signature is used to train a machine-learning algorithm for detecting abnormal behaviour or product during and after the process.
To detect the signature in time evolution of the profile of each sample a template matching technique is used. For this reason, this region is selected manually from one of the sample profiles and used as a template for further template matching operation. Normalized cross correlation is applied for template matching operation as implemented in Matlab® software package.
The signatures are further processed to reduce the dimensionality of the data. Principal component analysis can be used to reduce effectively the dimension of the space of the representation of date and to represent them with a more meaningful set of parameters. We will keep the first n most important components resulting from PCA to feed to the further anomaly detection algorithm. Practically n can be much smaller than the total number of components. The coefficients of the components of the samples along each PC create a distribution with a specific mean value and variance. These parameters are used to approximate the distribution with a Gaussian. After the training stage, each new sample will be compared with the calculated distribution: if the sample resides far from the centre of the Gaussian, given a user-defined threshold, or in other terms if the value of the distribution function fall below a specific threshold for the sample, it will be considered as an anomaly. The distribution has the following form (see equation 4).
Where Sigma’s are standard deviation for the coefficients of each PC.
In practice, to prevent numerical underflow, logarithm of the distribution function is used. Therefor we define the following quantity that measures deviation from normality as defined in equation 5.
For verification of the emissivity, temperature values at the scan-position were compared to the values taken from the gap between billet and pressing plate. The value, taken from the gap is assumed to show the correct value, since emissivity is close to 1 in the gap due to geometrical considerations. Because of nearly equal temperature values at the gap and at the scan-line, differing just for 10K, the emissivity must be close to 1 at the scan-line too.
Figure 34 shows the result of the FDM calculation for the cooling down progress. The mean value of 10 measurements, taken from the thermal sequences when the rolling process starts, is 800°C (calculated by camera-software with an emissivity set to 1). This value is quite close to the calculated temperature of 820°C. Adjusting the emissivity to 0.965 in the camera-software displays 820°C, which means, that the real value of the emissivity is 0.965. Due to uncertainty in duration for transfer of the billet from the furniture to the initial rolling position, and limited accuracy in temperature measurement, this result is believed to be reliable enough for further evaluation.
In addition it was found, that there is a lower surface temperature at two positions at the circumflex of the piece. It is caused by the contact of the working piece with the top and bottom rolling plates and consequently heat transfer to these colder objects. This is the explanation for the periodic pattern in the thermal sequences of the revolving part. In each revolution, two colder stripes are observed, as shown in all the time evolution of profiles (e.g. Figure 32).
After validation/calibration of the camera readings, several cross wedge rolling process are recorded in action for the purpose of training procedure. Template matching using normalized cross correlation is applied for finding the signatures in the time evolution of the profiles of the recorded videos. Fig. 6 shows the template and the automatically detected signatures of two other samples. Finding the match does not guaranty the existence of a normal signature in the profile. For example the matched signature in Figure 35 (right), that is related to a rolling process with a much faster rolling speed than the two other signatures, clearly shows a deviation from what it has been seen before.
Therefore, the match has to be further analyzed for anomaly detection. Since the signature is repetitive among different samples, dimensionality reduction techniques can reduce the number of parameters that are needed to represent the signature from thousands (that is the number of the pixels to represent each signature) to a small quantity of numbers. In this work, principal component analysis is performed for reducing the dimensionality of the samples.
Since the number of training profiles is limited, it is needed to produce more training examples from existing videos. This is achieved in two steps: 1- shifting the profile line (Figure 32) off axis, 2- shifting the signature in time, i.e. shifting the area of interest in Figure 33 to right and/or left. This way, about 120 signatures are created from each video resulting in about 1600 total training signatures each consisted of 3600 pixels. Producing these extra samples by such shifts is reasonable, since they can be physically observed due to fluctuations in measurement or the process.
After training and during the production process, any time a part is produced the video segment containing the rolling process is evaluated for detecting anomaly. The operations in this stage are performed using the results of the principal component analysis. The evaluation profile (presented initially as a vector of pixel values) is projected along the vectors of PCs and the resulting coefficients (xi) are used to calculate deviation (eq. 5).
The variance of the training data set along the nth PC as n increases monotonically decreases (Figure 36). In order to avoid numerical errors due to division to very small numbers in eq. 5, deviation will be calculated using the first 150 PCs. Based on the deviation values calculated for the samples that are known to be good samples, a threshold is defined. The samples with a deviation above the threshold will be flagged as abnormal.
Figure 37 shows the result of the calculations on three different sets. Set 1, contains the training data (without the previously mentioned shifts in space and time), set 2 is an evaluation set produced with the same rolling condition as for the training set, and set 3 is a set of samples produced with an abnormal rolling speed. As it is clear from the results all the samples that are produced by an abnormal parameter, have a deviation value higher than normal samples, except the sample labeled as S1, that as it is clear from the image contains abnormalities in the signature. Based on this plot a threshold value of 1000 seems to be a reasonable value for threshold (Fig. 8, red line).
Finite difference simulations and a machine learning method were used to set up an automated quality control system for cross wedge rolling process. Machine learning algorithm utilizes template matching, PCA and anomaly detection methods. Principal components analysis effectively reduced the number of presentation parameters to a small and more meaningful set of coefficients. In evaluation stage, a sample is flagged as abnormal if the projection of it on PCs results in coefficients that are unexpected compared to the distribution of the coefficients of the training set. The analysis could robustly distinguish between normal and abnormal samples.
The use of general tools such as thermography and mathematical models like template matching and PCA makes this method easily adoptable for various production processes that consist of a repetitive task producing a thermal signature.
The evaluation software, written as Matlab®-stand-alone executable, is now tested, improved and ready to be used for automated checking, when evaluation is running on a PC-platform. For future use, this software could be translated into other platforms to be embedded into CRW-machine-hardware.
Potential Impact:
In the following the planned dissemination activities are described. All partners approved it at the final steering committee held at OMTAS in Gebze, Turkey on 20th April 2016. Thereafter all partners had the chance to make comments on or to make their objections against the planned way of disseminating the results as it was send to everybody before the meeting. The planned actions below are agreed by all partners and will be performed within 1 year after the project duration. In the following, the planned dissemination activities and IPR are described for each SME and RTD partner.

IPH (RTD)
Distributing the CWR machine
In the recent past, Industry was more and more interested in CWR technology. CWR is a very resource efficient preforming technology and leads to a significant reduction of flash. But due to the high costs for CWR machine many companies do not use the CWR technology. Therefore IPH developed a CWR machine especially to the needs of SME’s. This machine is much cheaper and IPH expects to find customers which will be highly interested in the new developed CWR machine.
Presenting results on a conference
IPH presented the results at the International conference on recent trends in materials and mechanical engineering, 15th – 16th of January 2015, Auckland New Zealand. The main content was the comparison of simulations of cross wedge rolling processes with real trials.
Additionally IPH will present the CWR machine 2017 at the UKH in Hannover. The UKH is one of the most important conferences on forming technology in Germany. The presentation aims to show the CWR machine to worldwide users.
Publication in scientific and industrial journals
IPH wrote four publication which have been already submitted. The publications were about the development of the CWR processes and the CWR machine as well as the presentation of the project objectives. IPH wrote both, scientific and industrial oriented papers. Within 2016 IPH will publish three more papers. Two of them will be published in highly scientific journals whereas one aims to reach the german forging industry in order to present the project results.
Lectures for university
The results of the project have been implemented in the lecture “Umformtechnik – Grundlagen” “Umformtechnik – Maschinen” for the winter semester 2015/2016 as planned.
Project website
The project website www.covaform.eu will stay online for at least one year after the project. IPH will update the website with new publications made by the partners to show the results of the project CoVaForm.
Furthermore, the download area of www.covaform.eu will stay the safe platform for important internal data of the project like: General project issues, deliverables and data of project meetings.
PROFACTOR (RTD)
The process inspection is becoming more and more important for production processes. PROFACTOR has a great experience in this technology. The project results allow PROFACTOR to widen their field of interest to the forging technology.
Presenting results on a conference

PROFACTOR presented the results of the project in:
1) Gerhard Traxler and Amirreza Baghbanpourasl, Monitoring of the cross wedge rolling process by thermography, 13th International Workshop on Advanced Infrared Technology & Applications, Sept.29 – Oct.2 2015, Pisa, Italy, ISBN 978-88-7958-025-0, pp214
Presenting results in forging associations
Thermografieforum Eugendorf, 18.9.2015, Austria: „Prozessüberwachung beim Querkeilwalzen“ was skipped due to illness, but is planned to be presentated by PRO in Sept. 2016.
Lectures for university
The results of the project have been implemented in the lecture “Thermografie und Wärmefluss” for the summer semester 2016 in electrical engineering as planned.

TECNALIA (RTD)
TECNALIA aims to acquire new customers in the forging industry due to the new knowledge gained in the project regarding the development of CWR processes for stepped shafts.
Publication in scientific and industrial journals
TECNALIA plans to publish, in collaboration with other project partners, the project results in International Journals and specialized magazines.
Presenting results on a conference
TECNALIA will show the simulation results in the yearly FORGE Users Conference in Cannes (France).
Presenting results in forging associations
TECNALIA, as partner of SIFE, Spanish forging association, takes part in its Technical Committee, and it is planned to show the project results to all SIFE partners in its next meeting on June 17th 2016.
Lectures for university
TECNALIA is in close contact with MAI and TKNIKA for the possible ways of showing the CoVaForm project activities during their courses about process design and manufacturing.

OMTAS (SME)
OMTAS is the main user of the CWR machine build up within the project at OMTAS.
Promotion in associations
OMTAS will hold a presentation of the results of CoVaForm at the DÖVSADER (Turkish forging association). The paper and the presentation will be prepared together with IPH and it will present the flash reduced process chain.

OMS (SME)
OMS uses the process design of TECNALIA for their product.
Promotion in associations
OMS will hold a presentation of results of CoVaForm at the UNISA (Italian forging association). The paper and the presentation will be prepared together with IPH and it will present the CWR machine.

AURRENAK (SME)
Promotion in associations
Due to their new gained knowledge, AURRENAK will present their new competence to existing and new partners. They will try to start collaboration with existing forges, which does not have the possibility to design CWR tools.

ERATZ (SME)
Promotion in associations
ERATZ will hold a presentation of the results of CoVaForm at the IMU (German forging association) in the year 2017. The paper and the presentation will be prepared together with IPH and it will present the method to design CWR tools.

List of Websites:
www.covaform.eu

Related information

Documents and Publications

Contact

Bernd-Arno Behrens, (Managing partner)
Tel.: +49 511279760
E-mail
Record Number: 199519 / Last updated on: 2017-06-20
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top