Lasers in Eng., Vol. 18, pp. 23–33 Reprints available directly from the publisher Photocopying permitted by license only

©2008 Old City Publishing, Inc. Published by license under the OCP Science imprint, a member of the Old City Publishing Group

Optimization of Laser Cladding by Using a High Speed Process Frank Vollertsen and Knut Partes∗ BIAS – Bremer Institut fuer angewandte Strahltechnik, Klagenfurter Str. 2, 28359 Bremen, Germany

Laser cladding is often used as a repair process for tools and workpieces. The disadvantages of the state of the art are the low speed (typically 1 m/min) and the high powder consumption. In this study a strategy will be given how to optimize the process with respect to time, cost and quality simultaneously. Benchmark numbers for the effective processing speed, taking the necessity of a post treatment by grinding into account, were developed. The basic idea was the reduction of heat conduction losses by an increase in the processing speed by a factor of 10 or more. The results show that the quality can be enhanced by reduction of the waviness of the clad surface, whereas the powder losses and processing time are reduced also. The quantitative description of the enhancement is achieved by using benchmark numbers. Keywords: Optimization, Laser Cladding, Nd:YAG Laser, Powder Efficiency, Energetic Efficiency.

1 INTRODUCTION Laser cladding is a technique to generate protective layers on metallic surfaces or building up metallic structures [6]. It can be used to improve the corrosion resistance [8], wear resistance [7] of parts or as a layer for regeneration applications [1]. Competitive technologies are thermal spraying and conventional overlay welding. In contrast to thermal sprayed layers, laser clads are dense with almost no porosity and a dilution zone, which is important to form a strong bond to the base material [2]. In contrast, conventional overlay laser cladding is more precise in it’s deposition due to the accurate spatial limited energy distribution. Near net shaped structures can be produced by laser cladding. ∗ Corresponding author: E-mail: [email protected]

23

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 23 — #1

24

Frank Vollertsen and Knut Partes

The energy input is comparably small and the distortion of the part can be kept small, [3, 7]. However, the energy deposition can be optimized, [4]. Despite all these advantages laser cladding is an expensive technology [9]. In contrast to other processes much time is taken to treat an area. Another important aspect is the exploitation of the laser power and the powder. As a result of the increasing costs of energy and resources the impact of the efficiency is becoming increasingly more and more important. On the other hand, the demands on the quality are also increasing. A constant acceptably, quality laser clad today is usually too expensive to be used in industrial manufacturing. In this study a way is shown how to benchmark the laser cladding process in terms of efficiency, processing time and quality. It is necessary for the process to increase some parameters without decreasing the others. In the following a strategy is given how to improve the laser cladding process.

2 BENCHMARKING In order to benchmark a process in production engineering it is important to take various aspects into account. In quality management it is well known as the “magic triangle”. The “magic triangle” is an area spanned in a 3 dimensional coordinate system. The coordinates in the case of laser cladding can be represented by the costs, the production time and the complexity of post processing, see Figure 1. In order to improve a production process from the quality management point of view, the area of the triangle must be minimized. For example, it is often difficult to decrease the production time without increasing the costs or the complexity of post processing. The first step is to set up benchmark numbers for all of the three criteria. These benchmark numbers will be described in the following.

FIGURE 1 Magic triangle.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 24 — #2

Optimization of Laser Cladding by Using a High Speed Process

25

2.1 Costs In laser cladding processes filler material is melted onto surfaces. Hence, energy and material are necessary for laser cladding. The costs are represented by the energetic exploitation of the laser power and the powder efficiency. Therefore, an energetic efficiency and a powder efficiency is defined as the benchmark of the process. The powder efficiency is the ratio between the mass brought onto the specimen and the mass supplied to the process: m (1) m ˙ ·τ The mass difference of the specimen, m, is determined gravimetrically. The mass supplied to the process is given by the product of the powder feed rate m ˙ and the total processing time τ . The energetic efficiency ηenergy is the ratio of the energy, which is necessary to melt the filler material on the specimen and the energy supplied by the laser beam: h · m ηenergy = (2) P ·τ The energy necessary to melt the mass of filler material onto the specimen is given by the product of the difference in mass, m, and the specific constant h. h includes the energy to raise the temperature of the filler material from room temperature to the liquidus temperature and in addition, the melt enthalpy (Equation (3)). The energy supplied by the laser beam is given by the product of the laser power P and the total processing time τ . ηpowder =

h = cp · T + qs

(3)

The calculation of the melting energy via the specific energy term h is given by the product of the specific heat capacity cp and the temperature difference between the liquidus temperature and the room temperature T added to the specific melting enthalpy qs . Hence two benchmark numbers for the costs of a laser cladding process are given in order to improve the cladding process. 2.2 Production time The simplest definition of the production time is the total processing time. The processing time of two processes can be compared if the area and the geometrical conditions such as track offset are the same. Comparing the same cladding processes and treating the same areas in the same manner, the total production time is proportional to the inverse of the scanning speed. Another possibility to benchmark the production time is to measure the deposited mass per time M, analogue to the powder efficiency: M=

m τ

(4)

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 25 — #3

26

Frank Vollertsen and Knut Partes

FIGURE 2 Dimensions of a clad profile.

The value can be calculated by the measurement of the deposited mass during a process divided by the total processing time. 2.3 Complexity of post processing The complexity of post processing is an indirect representative number of the cladding quality. The criteria of the quality itself are dependent on different failure mechanisms, such as cracks, bonding errors, pores and high surface roughness. A density of these errors can be defined such as cracks per cross sectional area. Nevertheless an evaluation of the quality of a clad will only have binary results. The clad passes or fails. Clads are presumed to be free of cracks. The usable clad mass has to be determined in order to ascertain the complex nature of the post processing. In general, clads produced by laser cladding have to be subsequently machined in order to obtain a defined surface. Hence the clad height must be reduced to the minimum clad height, see Figure 2. The usable clad mass must be normalized as do the other quality numbers. The usable clad mass ηmass is calculated by: ηmass =

w · hmin · l · ρ m

(5)

The product of the width w, the length l and the minimal height hmin of the clad represent the usable volume multiplied by the density ρ of the clad material. The usable mass can be calculated. 2.4 Benchmark numbers In order to benchmark the process a compromise has to be done for getting all the benchmarking criteria together. The numbers should be dimensionless and normalized in the sense that the values are between 0 and 1. Every benchmark number will be divided by the maximum measured value. Thus, the best value is always one.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 26 — #4

Optimization of Laser Cladding by Using a High Speed Process

27

The benchmark number for the costs ηc can be seen from the point of the powder efficiency or from the energetic efficiency. Nevertheless, the product of the two efficiencies results in a combined efficiency. The combined efficiency takes powder and energetic efficiency in the same weighting into account: ηc =

ηenergy · ηpowder ηenergyMAX · ηpowderMAX

(6)

For the production time the mass brought onto the specimen per time will be taken into account. This value ηt is normalized as the other benchmark numbers: ηt =

M MMAX

(7)

The complexity of post processing will be represented by the usable clad mass. The usable clad mass is a dimensionless normalized number ηf as the benchmark numbers for the production time and the costs: ηf =

ηmass ηmassMAX

(8)

The overall benchmark number will be given by the area of the “magic triangle” by simple geometry assumptions. The area itself is only dependent on the three benchmark numbers. The overall area is normalized in the sense that the maximum area is one. The smallest area marks the best process quality. In (Equation 9) the overall benchmark number is given. In this study the overall benchmark number will be called  value. The smaller the  value the better the overall benchmark.
(1 − ηc )2 (1 − ηt )2 + (1 − ηc )2 (1 − ηf )2 + (1 − ηt )2 (1 − ηf )2 = 3 (9)

3 EXPERIMENTS The material combination of Stellite 21 as filler material and conventional steel as the base material was chosen for the experiments. The laser beam was generated by a solid state Nd:YAG laser with a maximum output power of 4 kW. The powder was supplied to the process via a pneumatic powder feeding device. The working heat consists of the laser focussing unit and an off axial nozzle. The base material was cylindrically shaped in order to obtain comparably high scan speeds by the rotation speed, see Figure 3. All cladding experiments were undertaken with a track offset of 0.5 mm treating the same area size. The width of the clad was 10 mm and the diameter of the specimen 30 mm.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 27 — #5

28

Frank Vollertsen and Knut Partes

FIGURE 3 Setup of the laser cladding experiments.

The approach in order to make a significant improvement to the process is to increase the scanning speed and use the effect of the reduction of the heat conduction losses [5]. The energy per unit length and the line per unit length are given in (Equation (10)) and (Equation (11)) and were constant for all the experiments. P (10) v m ˙ m= (11) v The line energy E is given by dividing the laser power P by the scanning speed. The mass per unit length is given by dividing the power feed rate by the scan speed. Another boundary condition was the process energy supplied by the laser beam. This value is given by the product of the laser power P and the overall processing time τ : E=

EProcess = P · τ

(12)

All experiments were undertaken above the conventional scanning speeds of 1 m/min with intermediate scanning speeds up to 21 m/min. In the experiments, parameters for m and E were found in order to produce defect-free clads over a large variation of the scanning speed. The empirically found values were: a line energy of 8 J/mm and a line mass of 2.3 g/m.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 28 — #6

Optimization of Laser Cladding by Using a High Speed Process

29

4 RESULTS 4.1 Costs The cost efficiency ηc increases with the scan speed. The benchmark number for the process costs increases with the process parameters, which are related to the boundary conditions. The maximum of the measured benchmark number is at the highest scanning speed of 21 m/min. 4.2 Processing time The mass per time ηt increases with increasing scanning speed. The mass per time varies from 0 to 1 between the process parameters. The highest mass per time was determined at 21 m/min, see Figure 5.

FIGURE 4 Cost benchmark number with respect to the scan speed.

FIGURE 5 The processing time benchmark number with respect to the scan speed.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 29 — #7

30

Frank Vollertsen and Knut Partes

FIGURE 6 Quality benchmark number with respect to the scan speed.

FIGURE 7 Overall benchmark number with respect to the scan speed.

4.3 Quality The effective mass ηf varies with the scanning speed. At scan speeds of 3.6 m/min and 6 m/min the clad was not continuously distributed over the surface. Therefore, the minimum clad height is 0 and hence the effective mass is also 0. The effective mass decreases with the increasing scan speed, Figure 6. 4.4 Overall benchmark The  value was calculated by (Equation (9)) and is shown in Figure 7 with respect to the scanning speed. It can be seen that the overall benchmark number decreases with increasing scan speed.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 30 — #8

Optimization of Laser Cladding by Using a High Speed Process

31

5 DISCUSSION The overall quality can be described by the  value. This value is influenced by three measured values. The costs are correlated to the cost efficiency, which increases markedly with the scan speed. It was shown in [3] that this is correlated to decreasing heat conduction losses and a more efficient coupling of the laser power into the powder particles connected to the boundary conditions. Nevertheless, with increasing efficiency the mass brought onto the specimen is also increasing whereas the processing time decreases at higher scan speeds. Hence, the mass per time increases also. The effective mass is zero at the lowest speed, because cladding does not cover the surface of the parts continuously. This produces a minimal clad height of zero for low scanning speeds. It can be seen that the effective mass decreases slightly with an increasing scanning speed, except for 18 m/min. A general decrease in the minimum clad height at very high scanning speeds can be caused by centrifugal forces on the melt pool produced by the rotation of the specimen.  increases with the scanning speed. This produces a much more overall efficient process with higher scan speed even with slightly increasing the post processing complexity towards higher scan speeds. This approach made a significant improvement of the parameters obtained.

6 SUMMARY In this study it was shown how to benchmark a production process such as laser cladding. The most important results are: - An overall benchmark number can be defined by taking into account the complexity of post processing, costs and processing speed simultaneously. - Sound clads can also be generated at scanning speeds, which are much higher than conventional scanning speeds. - The optimum quality – with respect to the total benchmark number – is derived at a scanning speed of 18 m/min, e.g. a factor of 15 above the conventional speed.

ACKNOWLEDGMENTS The authors would like to thank the DFG (Deutsche Forschungsgemeinschaft) for the financial support of this work (SE 226/52) in the priority program SPP1139.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 31 — #9

32

Frank Vollertsen and Knut Partes

Symbol

Unit

ηpowder m

g

m ˙ τ

g/min s J/g W J/(g K) K

qs M

J/g g/s mm mm mm g/mm3 g/s -

ηenergy h P cp T

ηmass w hmin l ρ

ηenergyMAX ηpowderMAX ηc MMAX ηt ηmassMAX ηf  E v m

EProcess

J/m m/min g/m J

Description Powder efficiency Mass brought onto the specimen during the process Powder feed rate Process duration time Energetic efficiency Enthalpy term Laser power Specific heat capacity Temperature difference between melting and room temperature Melt enthalpy Applied mass per time Usable clad mass fraction Width of the complete cladding Minimal height of the cladding Length of the coating Density of the cladding Maximal measured energetic efficiency Maximal measured powder efficiency Benchmark number for the costs Maximal measured mass per time Benchmark number for the production time Maximal measured usable clad mass fraction Benchmark number for the complexity of post processing Overall quality number Line energy Scan speed Line mass Process energy

REFERENCES [1] Vollertsen, F., Partes, K., Meijer, J. (2005). State of the art of Laser Hardening and Cladding. Proceding of WLT Conference: Lasers in Manufacturing, 281–306. [2] Ion J.C. (2005). Laser Processing of Engineering Materials. Butterworth-Heinemann Ltd. [3] Sexton, L., Lavin, S., Byrne, G., Kennedy, A. (2002). Laser cladding of aerospace materials. Journal of Materials Processing Technology, 122, 63–68. [4] Römer, G.R.B.E., Meijer, J. (200). Inverse calculation of power density for laser surface treatment, Annals of CIRP, 49(1), 135–138.

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 32 — #10

Optimization of Laser Cladding by Using a High Speed Process

33

[5] Partes, K., Seefeld, T, Sepold, G., Vollertsen, F. (2006). Increased Efficiency in Laser Cladding by Optimization of Beam Intensity and Travel Speed. Proc. Of SPIE vol. 6157, 42–52. [6] Steen, W. M. (2003). Laser material processing – an overview. Journal of Optics A: Pure and Applied Optics, 5, 3–7. [7] Sexton, L., Lavin, S., Byrne, G., Kennedy, A. (2002). Laser cladding of aerospace materials. Journal of Materials Processing Technology, 122, 63–68. [8] Jendrzejewski, R., Sliwinski, G., Conde, A., Damborenea, J. (2003). Laser cladding of Niand Co-based coatings for turbine industry application. Proc. of SPIE vol. 5229, 233–238. [9] Toyserkani, E., Khajepour, A., Corbin, S. (2005). Laser Cladding, CRC Press

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 33 — #11

“LIE” — “LIE11107” — 2008/2/29 — 12:01 — page 34 — #12