Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 108 (2022) 857–862
www.elsevier.com/locate/procedia

6th CIRP Conference on Surface Integrity

Monitoring the Surface Quality for Various Deep Rolling Processes –


Limits and Experimental Results
Oliver Maiß*, Karsten Röttger
ECOROLL AG Werkzeugtechnik, H.-H.-Warnke-Str. 8, 29227 Celle, Germany

* Corresponding author. Tel.: +49 5141-986551; fax: +49 5141-881440. E-mail address: oliver.maiss@ecoroll.de

Abstract

Mechanical surface treatments are specialized processes for improving part properties during their lifetime. However, they are additional
processes within the process chain, which generate costs and use additional resources. In order to reduce the reject rate within the production
line, process monitoring and process simulation need to be developed for these processes. Within the presented project, the effects of different
materials, pre-machining and deep rolling tools are analyzed to simulate the surface quality. The effect of process parameters on the surface
quality is measured to identify the limits of a surface integrity control system for deep rolling.
© 2022 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer review under the responsibility of the scientific committee of the 6th CIRP CSI 2022
Keywords: surface roughness, deep rolling, process monitoring

1. Introduction very flexible. For hydrostatic tools, the roller is always a ball,
which is pressed onto the surface by applying a hydraulic
Mechanical surface treatments (e. g. roller burnishing, pressure [6, 7, 8].
machine hammer peening, deep rolling or shot peening) are all The effect of the process on the surface roughness has been
focusing on the specific design of the surface integrity [1, 2]. analyzed in various studies in the past. Röttger analyzed the
Usually that means designing the near surface properties as effect of rolling parameters using hydrostatic tools for
residual stresses or hardness. These properties are affecting the hardened roller bearings [8]. The rolling force Fw, which is a
fatigue life of mechanically loaded parts [3, 4]. One often cited function of the rolling pressure p w, is affecting the surface
case is to increase the fatigue life of dynamically loaded parts roughness. Increasing the force, more plastic deformation
[5]. However, it is also well known, however, that all these occurs and the roughness decreases [8, 9, 10, 11]. Different
processes are creating specific surface properties and studies analyzed the effect of the rolling feed f w on the surface
roughness values [1]. The focus of this research is the roughness. It is found, that similar to turning, the roughness
roughness after roller burnishing and its predictability for increases for higher feed values [12]. In the latest research
different tools and parameters. projects the overlap factor u is used to describe the roughness.
Roller burnishing can be conducted with two different tool The overlap factor u describes the overlapping contact area of
types: mechanical- and hydrostatic tools. For mechanical tools, two rolling traces with the distance fw. It can be calculated by
the rolling force on the rolling element is applied mechanically
by using springs, whereas for hydraulic tools the rolling force 𝑢𝑢 = 1 −
𝑓𝑓𝑤𝑤
(1)
𝑏𝑏
is generated by hydraulic pressure. The rollers are rolling on a
support roller or bearings. In this case the roller geometry is

2212-8271 © 2022 The Authors. Published by Elsevier B.V.


This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer review under the responsibility of the scientific committee of the 6th CIRP CSI 2022
10.1016/j.procir.2022.05.199

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
858 Oliver Maiß et al. / Procedia CIRP 108 (2022) 857–862

, where b is length of the contact area calculated from the To include the surface roughness from the turning process
hertzian contact [9, 13]. Although the hertzian contact is into the experimental parameters (table 2), three different
limited to pure elastic deformation, the literature shows that the turning operations are used for the soft materials and three
hertzian contact can be used to explain simplified effects for process parameters for the hardened parts. The parameters are
deep rolling [14]. Maiss showed that the overlap factor is the chosen to achieve two similar roughness values with different
one independent parameter, to summarize rolling pressure, feeds and corner radii and two different roughness values with
rolling feed and ball size for hydrostatic tools rolling hardened just one feed value. Parameters for hard turning are chosen by
steel [9]. the same procedure.
Predicting the surface roughness is analyzed by using
analytical models as well as Finite Element simulations [15, 16, Table 1 Materials for the experiments.
17, 18]. Within the analytical models the hertzian contact is Nr. Material Condition Tool
used. Like the theoretical roughness in turning, the roughness 3.1645 AlCuMgPb 95 HB HG
can be calculated by using the simple formular 3.2315 AlMgSi1 85 HB EG

𝑓𝑓𝑤𝑤 1.0503 C45 247 HV5 HG


𝑅𝑅𝑡𝑡ℎ ≅ (2) 1.0503 C45 Hardened (40HRC) HG
8𝑟𝑟𝑏𝑏
1.7225 42CrMo4 347 HV5 HG, EG
, with rb as the radius of the roller ball. 1.2436 X210CrW12 292 HV5 HG
Loh et al. introduced this concept, but the results have not
1.3505 100Cr6 250 HV5 EG
been very promising [16]. A more complex model was
introduced by Hiegemann et al. for predicting the roughness for 1.3505 100Cr6 Hardened (50HRC) EG
rolling of coated surfaces [17]. Denkena et al. analyzed the
effect of rolling parameters for different materials and Table 2. Turning parameters for the pre-machining.
compared the results with the roughness value Rth [13, 19]. Name Corner radius r Feed f Theo.surf. roughn. Rth
To achieve specific machining results, the concept of VB1 0.4 mm 0.2 mm 12.5 µm
process control is well known in cutting operations [20, 21]. VB2 0.4 mm 0.36 mm 40.5 µm
Combining this with smart tools has become one trend in smart
VB3 1.2 mm 0.36 mm 13.5 µm
manufacturing (e. g. introduced SANDVIK a smart boring bar).
For roller burnishing process control has not been done quite For the first step a hydrostatic tool type HGx-9 from
often. A similar concept to SANDVIK has been introduced by ECOROLL AG Werkzeugtechnik is used with 5 different ball
ECOROLL to measure the rolling forces for mechanical rolling diameters. The process parameters rolling pressure p w and
tools, during the process [6]. overlap factor u is varied also in 5 steps. The parameters are
Using these data to measure and correct the process results summarized in Table 3. This leads to 25 different sets of
with the machine tool is the next step in the developments for process parameters which are applied on each material and
smart rolling tools. However, the user does not want to measure each pre-machining process. The parameters have been
the rolling force. Instead, the goal is to control the process adjusted in the case of aluminum; here the rolling pressure is
result, e. g. surface roughness or residual stress state. Therefore, reduced.
the scientific question of this research is, what are the limits of
Table 3. Experimental parameters for hydrostatic tool.
the prediction of surface roughness for roller burnishing, and
can the surface roughness be predicted by the rolling force, Name Unit Values
using a smart rolling tool? Rolling pressure pw [bar] 100; 150; 200; 250; 300
Tool [-] HG3; HG4; HG6; HG10, HG13
2. Experimental procedure Overlap factor u [%] 30; 45; 60; 75; 90

To achieve the described aim of this research, the For the second step, a mechanical rolling tool type EG14
experiments are conducted in three steps. The first step is to from ECOROLL is used. As in step 1, the process parameters
identify the main effects on the surface roughness for a simple rolling force Fw and rolling feed fw are varied in 5 steps, and the
contact geometry by using a ball with a hydrostatic deep rolling roller radius rR is varied in two steps, as shown in Table 4. The
tool (HG). In the second step the same procedure is applied for surface roughness is measured as in step 1.
a mechanical rolling tool (EG) with a more complex contact To identify the contact geometry and calculate the overlap
geometry. This step tests whether the hydrostatic tool findings factor u for the mechanical tool, a set of static penetrations are
can be applied for complex contact geometries. From these two conducted. The tool is pressed into the not rotating surface with
steps, the limits for a surface monitoring using the rolling force different forces. The resulting deformed surface defects are
will be identified in step 3. measured with a confocal microscope µSurf from Nanofokus
For the first and second step the deep rolling process is and analyzed using the software MountainsMap.
applied on different materials as summarized in table 1. For
each process, round parts are used, pre-machined by a turning
operation.

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
Oliver Maiß et al. / Procedia CIRP 108 (2022) 857–862 859

Table 4. Experimental parameters for mechanical tool. limitations. One limitation is the surface roughness of the
Name Unit Values previous operation. For low pressure values, the plastic
Rolling force Fw [N] 540; 1,070; 1,587; 2,143; 2,630 deformations are not enough to smoothen very rough surfaces,
as it can be seen in the middle section of figure 3.
Rolling feed fw [mm] 0.1; 0.32; 0.5; 0.71; 0.9
Roller radius rR [mm] 0.5; 2.5
Spring elongation x [mm] 0.1; 0.2; 0.3; 0.4; 0.5

In step 3 the mechanical tool has been equipped with the


ECOsense gauge from ECOROLL, which is a device to
measure the spring force within the mechanical rolling tool.
Measurement system and tool are calibrated to each other. For
the experiments, an EG5-40M tool is used (figure 1). The roller
has a diameter of 40 mm and a radius of 2.5 mm. As test
specimens, bars (1.7225) with a length of 200 mm and a
diameter of 45 mm are used. In the third step, the standard
deviation of the rolling force, and the roughness is identified
for 10 same processes. After this procedure the spring
elongation within the rolling tool x is decreased stepwise and
the rolling force and roughness values are measured.

Fig. 2. Surface roughness using HG-tool for various overlap factors.

Fig. 1. Experimental setup for using the smart rolling tool.

All roughness parameters are measured with a tactile surface


measurement system type MarSurf XR1 with a cut off
c = 0.8 mm.

3. Surface roughness for hydrostatic deep rolling

As already shown by Denkena et al., the surface roughness


for ball burnishing can be predicted by using a simplified
hertzian model [13]. Figure 2 shows the resulting surface
roughness for ball burnishing with a HG6 tool and a rolling
pressure of pw = 200 bar for all different pre-machining
operations and different materials. It also shows the predicted
roughness depending on the overlap factor u. It can be seen that
for overlap factors less then u = 60% the results fit quite well.
Also, there is almost no influence by the pre-machining.
In case of the variation of the ball diameter, there is an effect
of the pre-machining. Using a small diameter, the predicted and
measured values are drifting apart. A similar trend can be seen
for a small rolling pressure, shown in figure 3. Summarizing all
results from the first set of experiments using a HG tool, the
surface roughness can be predicted, however there are limits
regarding the pre-machining, the ball diameter, and the rolling
pressure. For large overlap factors, the model also has some Fig. 3. Surface roughness using the HG-tool for various rolling pressures.

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
860 Oliver Maiß et al. / Procedia CIRP 108 (2022) 857–862

4. Surface roughness for a complex contact geometry different than for the ball burnishing tool. The overlap factor
cannot be used as the most effective value to predict the surface
To identify the overlap factor, indentation tests are quality.
conducted. Figure 4 shows the different indents of the roller The reason for this is the asymmetric contact between the
with a radius of 2.5 mm into 100Cr6 (left) and AlMgSi1 (right) assumed indentations. Analyzing all experiments for the
and different forces. The length of these indentations has been mechanical tool, it is much harder to identify clear limits for
analyzed and correlated with the indentation during the force is the surface roughness prediction. With an increased force, the
applied. The graph in figure 4 shows the indentation depth for surface roughness decreases, but there are interactions with the
different materials. As it can be seen increasing the hardness, pre-machining and the material properties. To summarize these
the indentation decreases. From the pictures on the top, the experiments, general limits for the prediction, which can be
complex contact geometry can be seen. used within the next step of the experiments, cannot be
In contrast to the hydrostatic tools, where the contact identified. However, the measured roughness values can be
geometry is calculated from the elastic deformation using the used to analyze how the rolling force can be used as an
hertzian contact, the complex geometry can be calculated indicator for the roughness monitoring.
similarly. In a first step this research focuses on the general
approach of predicting surface roughness. Therefore, for the
complex geometry the plastic deformation is simply measured.
To compare both approaches, in the future the effect of plastic
or simple elastic deformation has to be analyzed in more detail.

Fig. 5. Surface roughness for mechanical tools by various overlap factors.

5. Limitation in monitoring the surface roughness

The tested process parameters in step 3 of the experiments


are typical parameters for this tool and material within the
industrial use of roller burnishing. The experiments for these
parameters show, that a mean surface roughness Rz = 1.35 µm
with a standard deviation of 3 = 0.153 µm. The standard
deviation is demonstrated in figure 6 with the grey shadow.
Analyzing the rolling force for these same processes shows,
Fig. 4. Static indentation of the roller into different materials. that the rolling force has a mean value of F w = 1,417 N with a
standard deviation of 3 = 224.1 N. The large difference
Analyzing the experimental results similar to the hydrostatic between the rolling force is due to the workpiece deflection.
experiments in step 1, similar graphs can be shown for the As it can be seen from figure 6 roughness value and force
mechanical tools. For example, for all pre-machining are leaving the shadowed area for the same process. This
operations and all materials, the rolling force influences the indicates that it might be possible to identify process errors
roughness. Increasing the force leads to smaller roughness regarding the surface roughness by using the rolling force.
values. However, compared to the ball burnishing, there are However, the results do not prove this concept in total. More
much higher effects by the material or the pre-machining. For experiments must be conducted.
ball burnishing, the surface roughness varies for the center of
the experimental plan by about 1 µm. It varies for the complex
contact geometry by about 10 µm.
Figure 5 summarizes the resulting surface quality for the
1.7225 steel and the pre-machining of Rz = 12.5 µm by
applying the overlap factor u. The graphs are much more

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
Oliver Maiß et al. / Procedia CIRP 108 (2022) 857–862 861

geometries and the increasement of the sensitivity in the force


measurement system.
In the future, similar projects must be conducted to apply
this approach to other surface and subsurface parameters
affected by deep rolling, e. g. residual stresses or hardness. The
presented research just focuses on the surface roughness. To
apply a similar approach for residual stresses would be great
future goal. Here the main issues are the predicting of the
residual stress state. Many projects focused on modeling
residual stresses however, they all were not able to transfer the
results to a different material heat treatment. General models
are not known so far and without conducting complex finite
element simulations, a approach as given in this paper, is not
possible. Research projects as from Meyer and Kämmler are
heading to this goal [14].

Acknowledgements

This project is funded by ECOROLL for its internal


Fig. 6. Comparison of roughness and force variation using the smart tool.
developments. The authors wish to thank our colleague Boris
Wostal. We also wish to thank the IFW Hannover for
6. Conclusion and Outlook conducting the Nanofokus measurements.
The experiments have shown that it might be possible to use References
the rolling force as an indicator to control the surface roughness
within the burnishing process. However, some steps must be [1] Schulze, V., Bleicher, F., Groche, P., Guo, Y.B., Pyun, Y.S. Surface
taken in the future to complete this goal. modification by machine hammer peening and burnishing. CIRP Annals –
In the first experimental setup it is shown that for ball Manufacturing Technology, Vol. 65, 2016, p. 809-832.
burnishing the contact area between the surface and the tool is [2] Scholtes, B., Vöhringer, O. Ursachen, Ermittlung und Bewertung von
Randschichtveränderungen durch Kugelstrahlen. Mat.-wiss. und
simple and can be controlled by the overlap factor which
Werkstofftech., Vol. 24, 1993, p. 421-431.
depends on the rolling feed and rolling force. For very rough [3] Pape, F., Maiss, O., Denkena, B., Poll, G. Enhancement of roller bearing
surfaces, small forces, or small tool diameters, the presented fatigue life by innovative production processes. Industrial Lubrication and
model by Denkena has some limitations. Similarly, for high Tribology, Vol. 71/8, 2019, p. 1003-1006.
overlap factors, the roughness reaches an optimum value which [4] Rodríguez, A., López de Lacalle, L.N., Celaya, A., Lamikiz, A., Albizuri,
J. Surface improvement of shafts by the deep ball-burnishing technique.
cannot be improved by increasing the overlap factor.
Surface and Coatings Technology, Vol. 206/11-12, 2021, p. 2817-2824.
In the second set of experiments, the transfer of the findings [5] Berstein, G., Fuchsbauer, B. Festwalzen und Schwingfestigkeit. Z.
for a more complex contact geometry are analyzed. In this case Werkstofftechnik, Vol. 13, 1982, p. 103-109.
a mechanical tool with its effect on the surface roughness is [6] Website ECOROLL AG Werkzeugtechnik,
used. The results show that due to the more complex contact https://www.ecoroll.de/produkte.html, 13.12.2021
[7] Website Baublies AG,
geometry a simple model cannot be proposed. However,
https://www.baublies.com/werkzeuge, 13.12.2021
general trends can be identified. As one example, decreasing [8] Röttger, K. Walzen hartgedrehter Oberflächen. Doctor-Thesis, RWTH
the feed leads to a smaller surface roughness. Limits are Aachen, 2003.
defined by the hardness of the specimen, the pre-machining [9] Maiss, O. Lebensdauererhöhung von Wälzlagern durch mechansiche
roughness, and the rolling force. Because there are general Bearbeitung. Doctor-Thesis, Leibniz Universität Hannover, 2019.
[10] Denkena, B., Grove, T., Maiss, O. Influence if the cutting edge radius on
trends in the results, it seems reasonable to work on a specific
surface integrity in hard turning of roller bearing inner rings. Production
roughness model for these complex contact geometries and to Engineering, Vol. 9, 2015, p. 299-305.
use the rolling force as an indicator for the resulting roughness. [11] Grzesik, W., Zak, K. Modification of surface finish produced by hard
In the third step of the experiments, the hardware ECOsense turning using superfinishing and burnishing operations. Journal of
is analyzed to detect roughness errors. Here some more Materials Processing Technology, Vol. 212, 2012, p. 315-322.
[12] Grzesik, W., Zak, K. Producing high quality hardened parts using
experiments have to be conducted in the future. ECOsense is a
sequential hard truning and ball burnishing operations. Precision
device to measure the rolling force during the process. The Engineering, Vol. 37, 2013, p. 849-855.
experiments show that roughness increase and force decrease [13] Denkena, B., Abrao, A., Krödel, A., Meyer, K. Analytic roughness
are correlating, and ECOsense can detect this trend. prediction by deep rolling. Production Engineering, Vol. 14, 2020, p. 345-
All in all, the conducted experiments show that it will be 354.
[14] Meyer, D., Kämmler, J. Surface Integrity of AISI 4140 after deep rolling
possible in the future to measure, to predict, and to correct the
with a varied external and internal loads. Procedia CIRP, Vol. 45, 2016, p.
process parameters for a specific roughness value after the 363-366.
roller burnishing process. Future work will focus on the [15] Yen, Y.C., Sartkulvanich, P., Altan, T. Finite Element Modeling of Roller
modelling of surface roughness values for complex contact Burnishing Process. CIRP Annals, Vol. 54/1, 2005, p. 237-240.

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
862 Oliver Maiß et al. / Procedia CIRP 108 (2022) 857–862

[16] Loh, N.H., Tam, S.C., Miyazawa, S. Investigations on the surface [20] Stemmer, S., Abdel, D., Schwenzer, M., Adams, O., Klocke, F. Model
roughness produced by ball burnishing. Int. Journal of Machine Tools and Predictive Control for Force Control in Milling. IFAC-PapersOnLine, Vol.
Manufacturing, Vol. 31/1, 1991, p. 75-81. 50/1, 2017, p. 15871-15876.
[17] Hiegemann, L., Wedding, C., Tekkaya, A.E. Analytical contact pressure [21] Denkena, B., Bergmann, B., Stoppel, D. Tool deflection compensation by
model for predicting surface roughness of ball burnished surfaces. Journal drive signal-based force reconstruction and process control. Procedia
of Materials Processing Technology, Vol. 232, 2016, p. 63-77. CIRP, Vol. 104, 2021, p. 571-575.
[18] Li, F.L., Xia, W., Zhou, Z.Y., Zhao, J., Tang, Z.Q. Analytical prediction
and experimental verification of surface roughness during the burnishing
process. Int. Journal of Machine Tools and Manufacture, Vol. 62, 2012, p.
67-75.
[19] Denkena, B., Krödel, A., Heikebrügge, S., Meyer, K., Pillkahn, P. Surface
topography after deep rolling with milling kinematics. Production
Engineering, Vol. 15, 2021, p. 587-593.

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.

You might also like