JTTEE5 24:63–74 DOI: 10.1007/s11666-014-0170-6 1059-9630/$19.00 Ó ASM International

Johannes Ko¨nig, Michael Lahres, and Oliver Methner (Submitted July 4, 2014; in revised form September 16, 2014) After 125 years of development in combustion engines, the attractiveness of these powerplants still gains a great deal of attention. The efficiency of engines has been increased continuously through numerous innovations during the last years. Especially in the field of motor engineering, consequent friction optimization leads to cost-effective fuel consumption advantages and a CO2 reduction. This is the motivation and adjusting lever of NANOSLIDEÒ from Mercedes-Benz. The twin wire arc-spraying process of the aluminum bore creates a thin, iron-carbon-alloyed coating which is surface-finished through honing. Due to the continuous development in engines, the coating strategies must be adapted in parallel to achieve a quality-conformed coating result. The most important factors to this end are the controlled indemnification of a minimal coating thickness and a homogeneous coating deposition of the complete bore. A specific system enables the measuring and adjusting of the part and the central plunging of the coating torch into the bore to achieve a homogeneous coating thickness. Before and after measurement of the bore diameter enables conclusions about the coating thickness. A software tool specifically developed for coating deposition can transfer this information to a model that predicts the coating deposition as a function of the coating strategy.

Keywords

deposition simulation, liner coating, optimization, spray distance, spray spot, twin wire arc spraying

Abbreviations

LDS

1. Introduction An actual study concerning energy loss in passenger cars reveals that only one fifth of the fuel consumption is used for propelling the car, whereby the most significant loss is located in the engine itself (Ref 1). The piston-linergroup is responsible for approximately 50% of the friction loss in an engine (Ref 2). This circumstance—reduction of the friction loss—was already focused on at Daimler AG in the last decade. The solution lies in using newly developed barrel technologies. To this end, two procedures were deemed especially practical for realizing the aluminum crankcase: casting in of the cast iron liners and thermal coating of the cylinder barrels (Ref 3, 4). As a result of the high deposition performance and efficiency, the minimal heat build-up temperature of the substrate, and the desired level of porosity, the TWAS method is one of the most cost effective of its kind (Ref 5). The LDSÒ

This article is an invited paper selected from presentations at the 2014 International Thermal Spray Conference, held May 21-23, 2014, in Barcelona, Spain, and has been expanded from the original presentation. Johannes Ko¨nig, Michael Lahres, and Oliver Methner, RD/RMP, Research & Development, Daimler AG, 89081 Ulm, Germany. Contact e-mail: [email protected].

Journal of Thermal Spray Technology

TWAS NEDC OFAT DoE CAE

Lichtbogendrahtspritzen (German word for ‘‘twin wire arc spraying’’) Twin wire arc spraying New European driving cycle One factor at time Design of experiments Computer-aided engineering

process [LDSÒ stands for ‘‘Lichtbogendrahtspritzen’’ (twin wire arc spraying)] was further developed to represent an internal coating process. This new process enables advantages in friction properties related to the combustion chamber of the engine in the cylinder liner. It leads to an improvement between the interaction of the piston-liner group and, furthermore, to a reduction in emissions. As a result of DaimlerÕs lightweight design strategy, the engine crankcase in passenger cars has changed from cast iron to aluminum during the last two decades (Ref 6). It is important to mention that aluminum is a critical friction partner for the piston rings (Ref 7). Therefore, cast iron as the cylinder liner material is used from case to case, depending on the demands of the single engine type. The disadvantage of a cast iron liner, however, is its substantial weight and size. This directly opposes the lightweight philosophy (Ref 8). One possibility is to substitute the cast iron liner with a deposited thin metal coating (by LDSÒ technology) consisting of an iron-carbon alloy which is finish machined in several honing steps. The process chain, consisting of a surface preparation, the coating of the

Volume 24(1-2) January 2015—63

Peer Reviewed

Quality Designed Twin Wire Arc Spraying of Aluminum Bores

Peer Reviewed

cylinder liner, and the final step of finish machining the coated surface, is called NANOSLIDEÒ—a trademark name since 2011. The advantages of this technology are evident: as a result of using optimized materials, the friction loss can be reduced significantly. The engine is reduced in weight and space is minimized, while maintaining the capacity of the crankcase. Wear of the liner and piston rings can be diminished through this step, which leads to positive CO2 emissions (Ref 9, 10). NANOSLIDEÒ impresses with high process stability and cost efficiency. Due to this fact, the internal coating technology of aluminum crankcases leads more and more into the focus of development activities at Mercedes-Benz since 2001. Based on scientific investigations, the twin wire arc technology was developed as the suitable technology for series production of cylinder liners (Ref 11). In 2005, Daimler, Inc. started as first to market the series LDSÒ process for crankcases in the AMG V8 gasoline engine (Ref 12). Based on that change, the complete system liner was continuously developed further to the NANOSLIDEÒ technology (Fig. 1). Currently, several engine families in the gasoline and diesel engine segments were fitted with the NANOSLIDEÒ technology. Even in engines with the highest proportion of power and cubic capacity in four-cylinder powerplants (e.g., M133-AMG), a fuel consumption of 6.9 l/100 km can be achieved in the New European Driving Cycle (NEDC). Worldwide, all emission guidelines such as the EURO 6 standard can be achieved with this highefficiency engine (Ref 13).

Fig. 1 Cylinder crankcase with NANOSLIDEÒ technology

coating thickness should be deposited to ensure that the process remains cost effective. This has a positive effect on tool life and tool wear of the following machining operations for the coating. Due to the high-volume ramp-up of different engine families with the NANOSLIDEÒ technology, fully automated and intelligent coating deposition equipment is unavoidable. On one transfer line, different engine types with different bore counts must be coated. Contrary to small batches and as a result of parallel ramp-ups of several engine families, the development time for coating strategies to produce quality-confirmed bore coatings must be greatly reduced.

2.2 Materials and Spraying Conditions

2. LDSÒ in Mass Production and Its Requirements 2.1 Motivation LDSÒ coating of cylinder liners is not possible with conventional systems as the torch must plunge into the cylinder. This is necessary to achieve optimal coating efficiency through almost a vertical coating jet in the liner (Ref 14). The objective of this process is to generate a characteristic bore that offers a generous, smooth surface. This surface structure can minimize the friction and facilitates sufficient lubrication properties. In comparison to the classic liner technology, no typical honing structures were embedded (Fig. 2a). Moreover, the characteristic pore structure is revealed, typical for the LDSÒ liner coating, which caters to an appropriate oil detention volume (Fig. 2b). In addition to the coating properties, geometrical aspects such as coating thickness above perimeter and height of the cylinder bore must follow defined constraints. A quality-confirmed machining of the as-deposited bore coating with a downstream machining process must generate a homogeneous coating deposition with sufficient coating thickness across the total bore. Only a minimum

64—Volume 24(1-2) January 2015

With conventional twin wire arc-spraying torches as described in DIN EN 657 (Ref 15), coating deposition for internal coating of cylinder bores cannot be optimally carried out. Therefore, the activities to develop a new torch generation started at the research and development department of Daimler AG in Ulm, Germany in 1998 (Fig. 3). The major element of the torch is patented wire and gas guidance. The wire material and the process gas are induced axially to the rotation axes of the torch. The rotation axis is parallel to the center line of the cylinder. The outlet direction of the two wires has a radial direction and an aligned direction. The advantage is that even with fluctuations in the wire supply, the arc is positioned in the gas jet every time. This enables an orthogonal spraying angle toward the substrate (Ref 16). In addition to the continuous further development of LDSÒ torch systems, the entire periphery of the equipment technique is focused on. A special challenge consists in devising tailor-made coating strategies to produce qualitycompliant coatings for cylinder liners. With regard to the mass production of different engine types, the development time and development cost are a key challenge. Therefore, LDSÒ prototype equipment (Fig. 4) was installed in the research and development department of Ulm. A close connection exists in operating with series

Journal of Thermal Spray Technology

Peer Reviewed

Fig. 2 Honing surface of (a) gray cast liner and (b) NANOSLIDEÒ

Fig. 3 Development steps of the LDSÒ-torch technology

production. The equipment consists of the aforementioned torch, which is part of the rotating internal coating system. Wire spools, gas and current devices, and measuring technique are located at the rotating system to observe process-relevant parameters. The workpiece is handled using an industrial robot with six-axis kinematics. The robot transfers the status to the control level and activates, in real time, the positioning of the workpiece. Together with the robotÕs rotating axis and an additional external rotating axis, significant variances exist for handling the workpiece and the interaction between coating torch and workpiece. Intelligent gas management and a specific current source exist for control and regulation of the process parameters. The equipment also features a central particle extraction system. This is separated into an active and passive section. The active extraction section is responsible for the removal of non-adhesion particles during spraying and draws through the crankshaft zone of the crankcase. In the passive extraction section, the particles are extracted outside of the crankcase. The volume flow of both extraction parts can be regulated independently via butterfly valves.

Journal of Thermal Spray Technology

Fig. 4 LDSÒ-prototype laboratory

The control level of the equipment is an industrial computer that is responsible for all measuring data delivered by a bus system. The availability of all parameters of the equipment enables the control unit for supervision in order to achieve the necessary results. The particle properties are dependent on the process input parameters and influence, in particular, the formation of the coating properties (Ref 5). The process parameters, defined by the LDSÒ technique, and the positioning of the rotated coating device to the workpiece, have a significant impact on the coating quality and coating thickness. Newbery et al. maintain that particulate characteristics such as speed and temperature very much hinge on the voltage, wire feed rate, and gas flow rate as input parameters (Ref 17–19). The particulate temperature and speed have an especially significant effect on the homogeneity of the layered structure, the porosity that is critical with respect to the cylinder barrel coating, and the coating thickness (Ref 20–24).A minor but important influence on the coating thickness is the exhaust strategy chosen. The distance between the location of the particle formation—the arc—and the location of the coating formation—the substrate surface—must be constant. This is

Volume 24(1-2) January 2015—65

Peer Reviewed Fig. 5 (a) Schematically illustration of the coating thickness problem and (b) voids after the finish honing process

necessary because the coating properties of the entire cylinder must be fitted between tight tolerances with respect to roundness and cylindrical form. In these investigations, the influence of an eccentric coating torch on the coating thickness formation is analyzed. A constant minimal coating thickness must be defined for the last machining operation of the bores through finish honing (Fig. 5a). It can be seen that in case of a local undercut of the minimal coating thickness, residual rough, as-deposited coating exists after finish honing. This is a great disadvantage in the finished state of the cylinder (Fig. 5b). In addition to the local undercut of the minimal coating thickness required, the global undercut is a further waste criterion for the following machining steps as voids (rough, as-deposited areas after finish honing) exist across the entire cylinder surface. To adjust the coating thickness at a constant porosity of the coating, the workpiece feed rate must be adapted. Attention must be paid to the cylinder head area (entrance or exit of the torch) and the lower area (crankshaft zone) because an ideal overlap of the focused spray jet through geometrical constraints is not possible. The workpiece feed rates must therefore also be fitted in those areas.

3. Experiments 3.1 Eccentric Coating Deposition In contrary to series production, the application of a high-precision and cost-intensive device is not possible for each prototype part. Due to the demanded flexibility of the parts handling, only an industrial robot can be used with a slightly adaptable setup system. This high flexibility in manufacturing has a disadvantage with a loss of accuracy. Therefore, the parts positioning aspect must be closely regarded.

66—Volume 24(1-2) January 2015

As a result of this fact, the consequence of eccentrically positioning the torch inside the cylinder is investigated for four coated bores. The difference to a central coating process is that eccentric positioning of the torch causes no overlap between the rotation axis and cylinder center line (Fig. 6). Regarding the eccentricity of the rotating axis of the torch in the X direction, the resultant coating distance is the lowest at the 270° position of the torch in comparison to the remaining axis positions. In addition, a shifting of the torch in the Y direction can diminish the coating distance. Such a displacement of the rotating axis of the torch has a significant impact on the properties of the particle beam and the resulting coating properties, as mentioned before. An expansion of the jet at a lesser coating distance is significantly lower because of the high divergence of the LDSÒ process (Ref 5). Additionally, the particleÕs impact focused more on the partÕs surface. To check the eccentric coating deposition, four cylinders with a displacement in the X direction of the rotating axis of the torch were coated. The cylinders were measured over each total cylinder liner on a coordinate measuring machine before and after the coating process to evaluate the coating thickness. In addition to identifying the cylinder diameter, the concentricity was measured since the coating thickness is important as a function of the cylinder angle. It is apparent due to the differently colored areas that the coating thickness varies with the cylinder angle (Fig. 7). The distance of the coating at 270° toward the center axis of the coated cylinder (X 0, Y 0) is less than the distance between 90° toward the center axis. To verify the measured values from the coordinate measuring machine, the coating thicknesses were additionally measured on a polished section at the cylinder angles of 0, 90, 180, and 270°. For each cylinder, ten pictures of the polished cross section and the average values were recorded. A significant change of 90 lm of the coating

Journal of Thermal Spray Technology

Peer Reviewed

Fig. 6 (a) Centrically coating and (b) eccentrically coating

with an unchanging feed rate and the coating thickness was averaged (Fig. 9). With an increase of the workpiece feed rate, a reduction of the produced coating thickness was realized. Unfortunately, no constant thickness values were observed. To achieve the required minimal coating thickness, additional iteration steps must be implemented, depending on the result. The homogeneity of the coating thickness across the total cylinder length needs several further local adaptations to the feed rate. Such iteration steps for prototype parts are very cost intensive and time consuming.

4. Optimization/Improvement 4.1 Correct Alignment

Fig. 7 Average coating thickness in a cylinder for an eccentric coating deposition (oversubscribed)

thickness is detected as a function of the coating distance (Fig. 8). The former results were confirmed. The results reveal that for a homogeneous coating deposition above the extent, an exact adjustment of the torch inside the coated cylinder is necessary.

3.2 Generating a Global Minimal Coating Thickness New coating strategies were developed for prototype engines by way of tests up until now. These strategies are based on experience gained for well-known part geometries. This was possible in the past because of a clear part spectrum. The conventional approach to determine the optimal workpiece feed rate for a required minimal coating thickness is to coat the part at first with a constant workpiece feed rate over the total cylinder length. The approach was executed with a six-cylinder prototype engine future series. Three cylinders were coated

Journal of Thermal Spray Technology

The reason for an inhomogeneous coating thickness is the unknown with regard to the exact position of the cylinder center axis. If the center axis was known, it could be aligned coincidentally to the rotation axis of the coating device. The coating torch could then plunge into the middle of the cylinder and the coating thickness would be as constant as possible over the extent of the cylinder bore. To correct the alignment of the cylinder center axis, the actual position of the workpiece coated respective of the malposition must be determined. For this determination of workpiece surfaces and their 3D positioning, the application of tactile measurement sensors as they are used in coordinate measuring machines is suitable. In the case of the LDSÒ prototype equipment at Ulm, the workpiece is handled by the industrial robot and the tactile measurement sensor is in a static position. The advantage of this procedure is the abolition of the workpiece transformation before coating to achieve a higher positioning accuracy. To adapt the measurement principle of modern machining centers, a switching probe LP2 with an HSI interface (Renishaw GmbH, Pliezhausen, Germany) was integrated in the existing prototype coating equipment (Fig. 10a). Signal processing on the part of the robot control plane is a time-critical operation. On the one hand, when detecting a positive sensor signal edge—probe has triggered—the system must store the actual shaft an-

Volume 24(1-2) January 2015—67

Peer Reviewed Fig. 8 (a) Coating thickness at different cylinder angles measured at polished cross sections and (b) polished cross sections of cylinder 2

To determine the exact position of the cylinder center axis by a given orthogonal alignment of the cylinder head joint face and the cylinder center axis, the horizontal alignment of the cylinder head joint face must be checked and verified. After the correction of a possible tangential deviation of the workpiece, the cylinder center axes must be determined. With a given rectangular alignment of all cylinder center axes, it is sufficient to determine the center point of each cylinder bore (Fig. 11). This ensures that the coating torch plunges into the middle of the cylinder (Ref 25). The high accuracy of the measurement system enables a correction of a possible incorrect position of the workpiece as well as a measured quantification of the coating thickness. The coated liners can be measured—after a cooling period to ambient temperature—directly at the LDSÒ prototype equipment (Fig. 10b). The advantage of this procedure is that the time required can be decreased substantially. Fig. 9 Variation of the coating thickness through workpiece feed rate

gles of the robot. On the other hand, the axes must stop as quickly as possible because of the limited angle of deflection of the probe.

68—Volume 24(1-2) January 2015

4.2 Development of a Coating Strategy 4.2.1 Empirical Approach. Before coating a specimen to identify the suitable parameters for the adjustment of the required coating thickness, it is necessary to determine the main effects with the highest influence on the coating thickness. For this reason, methods of statistical analysis

Journal of Thermal Spray Technology

Peer Reviewed

Fig. 10 (a) Measurement centering device mounted on LDSÒ flange and (b) during measuring process

Fig. 11 (a) Use of three points to describe the cylinder head joint face as a plane and (b) determination of the cylinder center point

are applied. Due to the simultaneous change of parameters in comparison to the change of only one parameter (one factor at time—OFAT) as it is verified in the linear regression analysis, the number of experiments to determine the interrelationship can be drastically reduced. The focus of this DoE (design of experiments) is based on the interrelationships between influencing values (the process input parameter voltage, gas, wire feed rate, rotation of torch, and workpiece feed rate) and the fixed variable (coating thickness). The results of the fullfactored design of experiments—the main effects are shown in Fig. 12. Due to the insignificant interactions between the parameters, it is sufficient to examine the key, main effects (Ref 26). It can be clearly seen that only two parameters—the wire feed rate and workpiece feed rate—show effects on the coating thickness, which are proved to be significant. Based on the fact that the coating properties such as porosity are affected by the particle properties, the only lever for adjustment of the

Journal of Thermal Spray Technology

coating thickness without significant cross-influence on other coating properties is the workpiece feed rate. With the aid of the identified parameter, it is possible to influence the coating thickness. As is the case with any empirical approach, it is now a question of gathering information through targeted, systematic experiments carried out on the workpiece. Figure 9 shows the respective coating thickness distribution along the workpiece height of three selected workpiece feed rates. Two things can be observed: A variation of the coating thickness can be achieved with the adjusting lever workpiece feed, and the thickness of the coating is not homogeneous using a constant feed rate. However, a homogenous coating thickness is required for the final machining steps. Since an appropriate parameter is available, the deviations from the required coating thickness can be corrected with local adjustments made to the workpiece feed rate. This practice, however, has some disadvantages. Further adjustments to the workpiece feed

Volume 24(1-2) January 2015—69

Peer Reviewed Fig. 12 Results of the DoE: main effects of coating thickness

rate, the corresponding positions, and of course another measurement of the produced coatings, go hand in hand. This, in turn, leads to an increase in cost and time expenditures. Furthermore, it is no trivial task. A more effective variation in terms of cost and time is promised by a non-empirical approach in which knowledge is gained without recourse to direct observation. 4.2.2 Numeric Approach. A scientific, methodical approach without going through the sometimes repetitive subsequence of ‘‘coating-measuring’’ represents the approach for computer-aided engineering (CAE) of the manufacturing process of a coated cylinder liner. With the help of numerical analysis, it can be performed using a model of the coating process, the input and target variables, and the manipulated variables, which are determined iteratively. Through support of a simulation tool to predict the coating thickness with respect to the input parameter, time and cost consumables of empirical approaches of coating strategy development will lead to an end. The aim of the simulation tool is to identify a set of coordinates and velocities to achieve the required homogeneous coating thickness (Fig. 13). The following input variables are necessary. On the one hand, the geometry of the liner must be provided for coating the workpiece. Additionally, strategies for different geometries of liners should be developed. On the other hand, the requirement of a homogeneous coating thickness distribution is a further input variable for the simulation tool. Furthermore, a variable is required for describing the deposition coat on a substrate. A flat test piece can therefore be used, which is then analyzed by a three-dimensional laser-based line scanner. The measured and post-processed height profile of the surface is an additional input variable. As shown in Fig. 8, no homogenous coating thickness using constant

70—Volume 24(1-2) January 2015

Fig. 13

Input and output variables of the simulation tool

feed rates exists. To accommodate the differences between the model and the reality, a deviation factor is also necessary to ensure a correct simulation. In determining the coating thickness respective of the deviation factor, a test coating with a constant feed rate is conducted. This can be done directly after coating, as described above.

4.3 Spray Spot Measurement System Special consideration must be given to the input variable spray spot. As a result of this input parameter, the simulation of the layered structure increases and decreases, and thus, also the calculation of positions and speeds for setting the target value—the coating thickness—appropriately. As mentioned in chapter 3.2.2, a 3D

Journal of Thermal Spray Technology

Peer Reviewed

Fig. 14 Optical spray spot measurement system

profile of the coating deposition is required for simulating the coating thickness. The first approach—the measurement of vertical and horizontal coating traces in the cylinder bore on the aluminum substrate using the coordinate measuring technology—is by definition a target-oriented approach. However, it is also time consuming and therefore the demand exists for another resource. To avoid this step, a system was developed for the optical measurement of spray spots on flat test piece (Fig. 14). The flat test piece is subjected to the particle beam for a defined time; a spray characteristic spot is based on the activated specimen. In order to prevent disturbing process variations in the start and final phase of the process and have those in the spray spot, a mask is used that makes it possible to shadow the flat test piece until the process is stable. Due to the fact that the optical sensor is a 2D line laser LLT2900100 (Micro-Epsilon, Ortenburg, Germany), the coated specimen must be moved translative using a stepper motor to determine the three-dimensional height profile of the spray spot.

5. Results

Fig. 15 Averaged coating thickness in a cylinder for centrically plunging of the torch (oversubscribed)

The average result of the four coated cylinders with a defined embedded mismatch and adjacent position-optimized coating is shown in Fig. 15. In comparison to the coating thickness distribution at the extent for an eccentric coating deposition, (Fig. 7), it can be seen that there are no significant differences in coating thicknesses for a coating deposition with centric plunging of the torch in spite of incorrect clamping of the workpiece. The indemnification of a centric plunging of the torch exhibits the basis for a quality-controlled cylinder coating deposition. This, in turn, enables the production of good parts. However, a further adjusting lever is fixed with the workpiece feed rate to achieve the coating thickness required. First of all, it must be checked whether the simulation results correspond with those of the experiments.

5.1 Central Plunging Coating Torch To produce a quality-defined coating of a cylinder liner, the torch must plunge centrically into the coated cylinder. The centering system used first measures the cylinder head surface, then defines the center points of the cylinder in one plane and finally calculates the cylinder center axis. The functionality of the algorithms and the hardware of the centering system developed were checked in the manner that a defined malposition of the coated workpiece was implemented consciously. As described in chapter 2.3.1 concerning the eccentric coating deposition, the mismatch in X direction was set up. However, prior to the coating process, the centering system was applied to the incorrectly clamped workpiece. Now it is ensured that through a detection and correction of the mismatch before starting the coating process, a centric plunging of the torch is given.

Journal of Thermal Spray Technology

5.2 Comparison of Simulated and Measured Coating Thickness To verify the simulation results, two flat test pieces were coated with different wire feeds of 7 and 9 m/min. The coating distance was chosen so that it corresponds to the distance of a real cylinder coating. The height profiles of the respective spray spots were measured with the spray spot measurement system and for the simulation prepared (Fig. 16). For the comparison with real measurements—the coating thicknesses—each of the two cylinders was coated with the same wire feeds as the flat test piece, and a constant component speed of 470 mm/min. was then measured in tactile fashion. As it has already been shown that at constant workpiece feed rate due to exhaust effects and short strokes, the coating cannot obtain a homogeneous coating thicknesses (see Fig. 9), these four-cylinder

Volume 24(1-2) January 2015—71

Peer Reviewed Fig. 16 (a) Conditioned spray spot with 7 m/min wire feed rate and (b) spray spot with 9 m/min wire feed rate

Fig. 17 Comparison of coating thicknesses made by simulation and real coating process with different wire feed rates

coatings were running without suction and with full coverage. It can be clearly seen that the simulated layer thicknesses and the measured thicknesses are very close (Fig. 17). Only in the region of 120 mm, the measured coating thicknesses are higher than those of the simulation. This can be attributed to the shutdown process of the coating. Until the coating device actually receives and processes the OFF signal, the coating process is applied in this area and increasing the coating thickness. This consideration is, however, only for verifying the simulation results. In the production of coating good parts, the change of the wire feed rate is due to a change in the particle characteristics. The simulation must be able to replicate the coating thickness that has been produced through close conditions.

72—Volume 24(1-2) January 2015

For this reason, the coating results at a constant coating feed rate of 470 mm/min. and a 7.7 m/min. wire feed rate under series production conditions (such as with an active exhaust and an original stroke) were regarded as the comparison level (see Fig. 9). In the first step, the coated flat probe was measured with the spray spot measurement system to generate a distribution function that enables the simulation of a basis for determination of the coating deposition. Furthermore, the deviation factor was determined, which leads, at constant feed, to an inhomogeneous coating thickness across the entire cylinder length (Fig. 18). The simulation tool generates, with the required geometries, a coating thickness distribution (Fig. 18) that almost entirely matches with the measuring results in Fig. 9. The simulation calculates a forecasted-coating thickness in the lower cylinder area of approximately 250 lm, equivalent to the measured coating thickness. In both cases, a maximum coating thickness of approximately 320 lm is given. Based on the positive adjustment of simulation and measuring results, kinematic parameters such as velocity and point coordinates were generated along the workpiece height with this simulation tool (Fig. 19b). The coating thickness requirement of 300 lm as an additional input parameter of the simulation tool offers a coating strategy that creates a quality-confirmed coating thickness distribution for the coating process. The dotted line shows the position-dependent movement profile above the cylinder height. The continuous line shows the measured coating thickness distribution plotted above the cylinder height. In Fig. 19(b), it is clearly evident that a significant improvement in the coating thickness distribution can be achieved through the calculated moving profile from the simulation. The coating thickness decrease at the border of the cylinder—when coating with a constant workpiece

Journal of Thermal Spray Technology

Peer Reviewed Fig. 18 Simulated coating thickness using the deviation factor; (a) 3D-view of simulated coated liner and (b) coating thickness on surface line

Fig. 19 Tactile measurements of coated good parts with (a) constant workpiece feed rate and (b) with optimized kinematic parameters using simulation tool

feed rate (Fig. 19a)—was eliminated. The complete peakto-peak difference was also reduced significantly. An offline determination of the coating strategies is possible by combining both systems. This leads to a significant reduction in the capacities required, both with respect to development time and prototype parts. Finally, the strategies developed can be transferred to series production.

6. Summary/Conclusions A system was presented that acts proactively to the deficient positioning of the torch through workpiece

Journal of Thermal Spray Technology

measurement and correction of the position. The high accuracy of actuating elements and sensor technology enables measuring of the coating thickness directly at the LDSÒ coating equipment. This system can be completed with a simulation tool that devises coating strategies on the basis of defined input parameters. In this way, a required homogeneous coating thickness distribution on the cylinder liner can be generated. By coupling the input and output parameters in a closely coordinated fashion, it is possible to develop flexible, expedited, and cost-efficient coating strategies for a quality-defined deposition of LDSÒ-coated cylinder liners.

Volume 24(1-2) January 2015—73

Peer Reviewed

References 1. K. Holmberg, P. Andersson, and A. Erdemir, Global Energy Consumption Due to Friction in Passenger Cars, Tribol. Int., 2012, 47, p 221-234 2. J. Schommers, G. Doll, R. Weller, T. Behr, H. Scheib, M. Lo¨ffler, and J. Bo¨hm, Optimizing Friction: The Basis for Safeguarding the Future of Combustion Engines, 33rd International Vienna Motor Symposium, Fortschritt-Berichte VDI, Reihe, 2012 3. S. Beer, Aluminium-Motorblo¨cke: Konstruktionen, Werkstoffe, Giessverfahren und Zylinderlauffla¨chen-Technologien fu¨r Leichtbau-Pkw-Motoren, Landsberg/Lech: Verl. Moderne Industrie, 2005 4. B. Gand, Beschichtung von Zylinderlauffla¨chen in AluminiumKurbelgeha¨usen, MTZ Motortech Z, 2011, 72(2), p 128-133 5. J.V. Heberlein, Thermal Spray Fundamentals: From Powder to Part, Springer, New York, 2014 6. H. Fuchs and M. Wappelhorst, Leichtmetallwerkstoffe fu¨r hochbelastete Motorblo¨cke und Zylinderko¨pfe, Motortechnische Zeitschrift, 2003, 64(10), p 868-875 7. K. Bobzin, F. Ernst, J. Zwick, T. Schlaefer, D. Cook, K. Nassenstein, A. Schwenk, F. Schreiber, T. Wenz, G. Flores, and M. Hahn, Coating Bores of Light Metal Engine Blocks with a Nanocomposite Material using the Plasma Transferred Wire Arc Thermal Spray Process, J. Therm. Spray Technol., 2008, 17(3), p 344-351 8. E. Ko¨hler, S. Beer, C. Klimesch, J. Niehues, and B. Sommer, Leichtbau beim Zylinderkurbelgeha¨use fu¨r aktuelle und zuku¨nftige Anforderungen, Motortechnische Zeitschrift, 2009, 70(10), p 712-721 9. M. Hahn and A. Fischer, Characterization of Thermal Spray Coatings for Cylinder Running Surfaces of Diesel Engines, J. Therm. Spray Technol., 2010, 19(5), p 866-872 10. K. Holdik, M. Hartweg, M. Michel, T. Behr, R. De Zolt, J. Bo¨hm, and F. Spennemann, 125 Years of the Automobile-DaimlerÕs LDS Cylinder Track Opens up a New Chapter in Engine Construction, Bd. 1, p 46-50, 2011 11. A. Heuberger, P. Izquierdo, T. Haug, T. Wittrowski, and F. Lampmann, Twin Wire Arc Spraying as a New Coating Technology for Liner-Free Cylinder Bores, Weld. Cut., 2004, 56(6), p 356-361 12. F. Eichler, A. Fu¨rschuss, M. Hart, R. Schaich, B. Tschamon, R. Illenberger, M. Glose, and W. Zimmermann, Der Antriebsstrang des neuen C63 AMG, 29th International Vienna Motor Symposium, Fortschritt-Berichte VDI, Reihe, 2008

74—Volume 24(1-2) January 2015

13. J. Gindele, T. Ramsteiner, J. Fischer, and B. Tschamon, The New 2.0-l High-Performance Four-Cylinder Engine from MercedesAMG, MTZ Worldwide, 2013, 74(9), p 26-33 14. W. Tillmann and B. Krebs, Influence of Handling Parameters on Coating Characteristics in Order to Produce Near-Net-Shape Wear Resistant Coatings, J. Therm. Spray Technol., 2012, 21(3– 4), p 644-650 15. E. DIN, 657, Thermisches Spritzen—Begriffe, Einteilung, 2005 16. D. Nowotni, C. Wanke, T. Haug, and P. Izquierdo, Internal burner, 2002 17. A. Newbery, P. Grant, and R. Neiser, The Velocity and Temperature of Steel Droplets During Electric Arc Spraying, Surf. Coat. Technol., 2005, 195(1), p 91-101 18. A. Pourmousa, J. Mostaghimi, A. Abedini, and S. Chandra, Particle Size Distribution in a Wire-Arc Spraying System, J. Therm. Spray Technol., 2005, 14(4), p 502-510 19. D. Hale, W. Swank, and D. Haggard, In-Flight Particle Measurements of Twin Wire Electric Arc Sprayed Aluminum, J. Therm. Spray Technol., 1998, 7(1), p 58-63 20. S.L. Toma, C. Bejinariu, R. Baciu, and S. Radu, The Effect of Frontal Nozzle Geometry and Gas Pressure on the Steel Coating Properties Obtained by Wire Arc Spraying, Surf. Coat. Technol., 2013, 220, p 266-270 21. W. Tillmann, E. Vogli, and M. Abdulgader, The Correlation Between the Coating Quality and the Moving Direction of the Twin Wire Arc Spraying Gun, J. Therm. Spray Technol., 2010, 19(1–2), p 409-421 22. N. Hussary and J. Heberlein, Effect of System Parameters on Metal Breakup and Particle Formation in the Wire Arc Spray Process, J. Therm. Spray Technol., 2007, 16(1), p 140-152 23. H. Liao, Y. Zhu, R. Bolot, C. Coddet, and S. Ma, Size Distribution of Particles from Individual Wires and the Effects of Nozzle Geometry in Twin Wire Arc Spraying, Surf. Coat. Technol., 2005, 200(7), p 2123-2130 24. M. Planche, H. Liao, and C. Coddet, Relationships Between InFlight Particle Characteristics and Coating Microstructure with a Twin Wire Arc Spray Process and Different Working Conditions, Surf. Coat. Technol., 2004, 182(2–3), p 215-226 25. J. Ko¨nig, M. Lahres, O. Methner, B. Wielage, and C. Rupprecht, Methoden zur exakten Positionierung von Beschichtungsaggregaten fu¨r eine qualita¨tsgerechte Beschichtung von Zylinderlaufbahnen, Werkstoffe und Werkstofftechnische Anwendungen, 2013 26. B.J.J. Wappis, Taschenbuch Null-Fehler-Management, Hanser Fachbuchverlag, 2010

Journal of Thermal Spray Technology