TECHNICAL ARTICLE:

Introduction

In the field of tube production research is regularly conducted to gain knowledge and optimise the process parameters, a necessity resulting from the strong competition and fast evolution of the market. The aim of this work is to design and to improve an FEM model able to correctly define the parameters affecting the hole formation and propagation during the rotary tube piercing process (Mannesmann process).

The software used to perform the simulations was a commercial FEM code, namely Deform 3D. In particular the three-dimensional model developed is able to calculate stress and deformation distributions into the round bar. A high degree of interesting information is obtained from the simulations performed and the most important results are shown in this article.

Pietra SpA Company, a partner in this research, has supplied the technological data introduced in the simulation model. A comparison between the experimental and FEM results shows that the FEM software [1, 2] is able to correctly represent the phenomena, so it is possible to use it to optimise the production process of rotary tube piecing.

Research and Production

For seamless tube production processes such as Mannesmann or Stieffel, the analysis of hole formation mechanisms is one of the most important topics to take into account. This approach has been studied by several experts in the tube production field, whose objective it is to discover new solutions that can improve the process technology [3-5].

In this article the results obtained by the research team of the University of Brescia in cooperation with Pietra SpA highlighted. This conducted study focused on the stress analysis into the round bar during the hole formation and its successive growth in rotary tube piercing process. In particular, the most important phenomenon that takes place during the process is the hole formation, especially where and when it starts to form. In fact, experimental evidence has shown that the hole starts before it reaches the plug position and that, if the hole begins to form too early, the inner material can oxidate. This actively determines a low quality of tubes (tubes must be scrapped) while if it starts too late (too near to the plug) the plug is subjected to an accelerated wear.

Figure1: The longitudinal section of the tube worked without the use of the internal plug

The main goal of the present research is to find a tool able to forecast when the hole starts to form (and why), and also how this process and geometrical parameters influence the hole formation mechanism. Some 2D studies have already been conducted on this topic [6-9]. Even if they are a rough simplification of the actual process, they must be seen as a rapid tool able to furnish fast suggestions in studying and optimising the process. On the basis of the 2D results, some 3D investigations can be performed within a limited parameter range. This is very useful when the time required to perform 3D analysis is considered.

Other researches [10] have studied a 3D model of the process, but the most evident limitation in this study was the fact that the hole formation was due only to the plug because the software is not able to break the material. This rendered the process similar to an extrusion rather than a rotary piercing process. Experiments conducted without using the plug have shown that the hole does always form, even if it is not calibrated (Figure 1).

The results reached with the present study are very important for the company and also for the following researches. In fact, all information obtained from the simulations with the knowledge of the Mannesmann process, can furnish several elements to understand the mechanism that allows the hole formation into the round bar and to help the process optimisation. The attention has been mainly focused on stress and strain distributions inside the tube and on the comparison of FEM results with experiments.

Tube Production Using the Mannesmann Process

Several different technologies allow tube production with or without welding. The most important are the extrusion process, the rotary tube piercing process and the welding tube process. Among the techniques used to produce seamless tube the most competitive – in terms of production rate and quality level – is the rotary tube piercing process.

Analysing the rotary tube piercing process it is possible to observe that the cause of hole formation is the double action (compression and rotation) generated by the rolling mill. In particular (as shown in figure 2), the roll shape – two semi cones joined along their base – together with the rotation of the rolls around the skew axis, acts to generate compressive forces on the round bar distributed on the contact surface. This causes the round bar rotation and translation.

Figure 2: Mannesmann process

As a consequence, friction forces are generated between rolls and round bar (Figure 3). The friction force value is calculated by the following formulas:

The friction force acts on the contact area between rolls and round bar. The force direction is orthogonal to the axis of rotation of the roll.

These forces have two components along the tangential and longitudinal directions of the round bar. The effects of these components are:

• A couple of forces: FAt and FBt, which determine the bar rotation
• Two parallel forces with the same direction and intensity: FAa and FBa, which determine the bar translation

Figure 3: Forces acting between rolls and round bar

If the transversal section of the bar is considered, the effect of the Mannesmann process is an increase in the compression ( ) and tensile stresses ( ) from the external surfaces of the round bar to the bar axis, as figure 4 clearly shows.

Figure 4: Stress distribution into a transversal section of the round bar

Figure 5: Rotary tube piercing process

The 3D Model Definition and Implementation

Starting from the technical drawings of the rotary tube piercing mill system installed in Pietra SpA – a company that produces seamless tubes by means of the Mannesmann process – all the system components have been modelled by means of a three dimensional CAD. So a 3D model of the system has been obtained and used to perform the simulations using Deform 3D. In particular, in order to reduce the computational time, only half of the rolls (first cone), which compresses the round bar, has been modelled. For the same reason the roll surface is considered smooth without knurl but friction between rolls and bar was set very high (m=0,85).

Figures 6 shows the geometry of the round bar, of the rolls and of the rolling guide (Diescher guide). As far as the plug is concerned, the attention is focused only on the nose area, because this is the zone where the hole begins to form.

Figure 6: The 3D FEM model

The process data has been defined in cooperation with Pietra SpA. The round bar diameter is 150mm; the material is AISI 1020, and the working temperature is 1250°C. The rolls velocity is 80rpm, their maximum diameter is 780mm and the minimum distance between the rolls is 135mm. The simulation starts when the round bar comes into contact with the rolls, and the operation is modelled as isothermal.

The round bar material is modelled as plastic material, while the plastic flow stress curves have been obtained from literature [12, 13]. All the other elements are considered rigid objects. The friction between the round bar and rolls has been considered constant
() and the friction factor (m) has been set equal to 0,85. The friction under consideration is very high, due to the fact that this process is performed at a high temperature. It is very difficult to lubricate the piece and the roll surfaces are knurled (as previously mentioned).

The simulations conducted have furnished the stress and strain distributions in the round bar during the rotary tube piercing process.

Results Analysis

Model Validation

The model has been validated by a comparison among the simulation results, the experimental results and the data reported in literature.

Figure 7 shows the FEM stress distribution and the micrography of a transversal section of the round bar, obtained after an in-process stop. The fibre distribution allows the identification of the compression and traction areas into the section. The match between the simulation results is very good and the model can be considered validated. As a consequence it is also possible to use the FEM software to forecast the stress distribution during the rotary tube piercing process under different working conditions.

Figure 7: Transversal section FEM results Vs macrography of the round bar

Hole Formation Analysis

Once the FEM model has been validated the attention has been focused on the study of hole formation and growing mechanisms. Figure 8 illustrates longitudinal section of the round bar. It is possible to identify the starting point of the crack and the growth of the hole along the piercing direction. The round bar comes from Pietra SpA, after an in-process stop during the rotary tube piercing process.

As stated for figure 5, in the experimental results reported in figure 8 the plug is located in an incorrect working position. This shows that the crack initiation is not influenced by the top of the plug but by the stresses generated by the rolls on the round bar (see also figure 1). The interesting zone is the area before the plug top.

Figure 8: Longitudinal section of the round bar – with start and growth of the hole

As far as the hole formation and its propagation are concerned, the simulation results have been compared with the experimental evidence by analysing transversal sections of the round bar. Figure 9 shows the agreement between experimental and simulation results in terms of maximum stress distribution and hole formation and growth. In particular the stresses are concentrated on the rolling axis and their value s increased along the piercing direction as the distance between the rolls is decreasing.

Figure 9: Hole formation on the round bar axis in the piercing direction before the plug, comparison between simulation ( : 0÷77Mpa) and experimental results

The right position of the plug is beyond the area of maximum section of the rolls (figure 10a). In this position the top of the plug is able to help the material breakage during the process where the limit value of the stress is reached. This is very important in order to obtain internal surfaces of the tube without oxidation and a correct wear of the plug.

Figure 10: Positions of the plug (a: right; b: behind, risk of oxidation; c: ahead, excessive wear)

In fact, if the plug is positioned behind the right position (figure 10b), the inner material can oxidate determining a low tube quality (tubes must be scrapped). However if the plug is too far ahead of the right position (Figure 10 c), it is subjected to an accelerated wear (figure 11).

Figure 11: Accelerated plug wear

Longitudinal Stress Analysis

Using the 3D model it is possible to analyse the stress distribution in all the three space directions, therefore providing better knowledge of the mechanism causing the material breakage. A detailed analysis of the stress distribution in a longitudinal section (figure 12) has been carried out to define the amount of the error that the 2D FEM simulation can introduce.

Figure 12: Axial stresses

The value of stresses acting along the longitudinal section remains constant and lower than the stress values in the transversal section (less than 10 per cent of transversal stresses). From this it is possible to state that axial stress can be neglected and so 2D simulations can be used to study the hole forming phenomenon, without introducing relevant errors.

This is very important, because to perform 2D simulation is faster than to perform 3D simulation. In fact, 2D FEM simulations require a small number of elements, so both the computational time and the database dimensions are reduced. So it is possible to use 2D simulations to investigate how the geometry and the technological parameters influence the hole formation and, most of all, where and why the hole starts to form [10-13].

Conclusions and Future Works

The results obtained establish that the software used for the simulations is able to forecast the stress distribution into the round bar before the contact with the plug. In this first study the simulation was stopped before the contact between the bar and the plug. This choice is due to the fact that the 3D FEM software used is not able to generate the material breakage. In fact to continue the simulation without material fracture could determine a round bar deformed by the plug, so simulating a process which is similar to an extrusion rather than a rotary piercing.

It is very important to highlight that the performed simulations validate Deform 3D as software able to study the Mannesmann process. In fact, it is possible to analyse the influence of several parameters on the stress and strain distributions and so to be aware of the final product quality.

The results obtained with this research are very interesting and it is highly important to develop in detail the FEM investigation of the rotary tube piercing process, with the aim of obtaining a simulative model as reliable and as close as possible to the actual process.

In future, the analysis will be focused on all those variables affecting the hole formation mechanism, including:

• Heat transfer
• Friction
• Roll geometries (diameter and skew angle)
• Material characterisation at high working temperatures
• Position of the top of the plug

The cooperation between the research and the production, realised due to the support of Pietra SpA, allows provides increased knowledge about the Mannesmann process. The interest on this topic is also confirmed by the new cooperation of the team of the University of Brescia with the Dalmine Group.

Acknowledgments

This work has been made possible with the cooperation of Pietra SpA Company, Italy

References

1. Mori, K and Osakada, K, ‘Finite Element Simulation of Three-Dimensional Deformation in Shape Rolling’, Int J Numer Mech. Eng, 30-8, pp. 1431-1440 (1990).
2. Osakada, K, Nakano, J and Mori, K, ‘Finite Element Method for Rigid-Plastic Analysis of Metal Forming – Formulation of Finite Deformation’, Int J Mech. Sci, 24-8, pp. 459-468 (1982).
3. Montelatici, L, Scartabelli, B, Tosato, S, Wolter, D, Dvorkin, EN, ‘Full-scale Experimental Determination of the Material Deformation in a Piercing Mill on a Seamless Tubes Rolling Line’, Proceedings of the 14th Rolling conference, IAS, San Nicolàs, Argentina (November 2002).
4. Cavaliere, MA, Berazategui, D, Goldschmit, MB, Dvorkin, EN, Montelatici, L, ‘Modeling of the Piercing Process: Preliminary Results’, 14 conferencia de laminacion y 4° Encuentro de la Secciòn Argentina de la Iron and Steel Socety, Buenos Aires, Argentina (2002).
5. Cavaliere, MA, Goldschmit, MB, Dvorkin, EN, ‘Finite Element Simulation of the Steel Plates Hot Rolling Process’, Int J Numrical Methods in Engrg, 52, pp. 1411-1430 (2001).
6. Ceretti, E, Giardini, C, Attanasio, A, ‘Analysis of Rotary Tube Piercing Process: Simulation and Experimental Results’, AITEM 01, Bari, Italy (September 2001).
7 Capoferri, G, Ceretti, E, Giardini, C, Attanasio, A, Brisotto, F, ‘FEM Analysis of Rotary Tube Piercing Process’ Tube & Pipe Technology, pp. 55-58 (May/June 2002).
8. Capoferri, G, Ceretti, E, Giardini, C, Attanasio, A, Brisotto, F, ‘Studio del Processo di Laminazione Obliqua agli Elementi Finiti: Simulazioni 2D- 3D e Risultati Sperimentali’, La metallurgia Italiana, pp. 60-63 (October 2002).
9. Ceretti, E, Giardini, C, Attanasio, A, Brisotto, F, ‘Further Improvements in Modelling a Tube Rotary Piercing Process’, VII AITEM, Cassino, Italy (September. 2003).
10 Pietsch, J, Thieven, P, ‘FEM Simulation of the Rotary Tube Piercing Process’, MPT International, pp. 52-60 (May 2003).
11 Voswinckel, G, ‘Developments in the Field of Piercing Billets for Seamless Tubemaking’, internal report of Mannesman.
12 Shirakashi, T et al., ‘Flow Stress of Low Carbon Steel at High Temperature and Strain Rate’, Bullettin of JSPE, Vol 17, pp.167-172 (1983).
13 Cavaliere, MA, Gomez, G, Gazzarri, JI, Perez, T, Dvorkin, EN, ‘Experimental Procedure for Determining True Stress-True Strain Curves for Steels in the High Temperature Range and Under Controlled Deformation Rate’, ECCOMAS 2000 (European Congress on Computationa l Methods in Applied Sciences and Engineering), Barcellona, CIMNE (2000).

University of Brescia
Department of Mechanical Engineering, Brescia, Italy
Fax: +39 030 370 2448
E-mail: ceretti@ing.unibs.it