Research and Application of Robot Force Control for Friction Stir

Welding

Yanggui Xin

, Ye Huang

, Yunqiang Zhao

1, a, *

, Shiyi Gao

and Su Li

Guangdong Welding Institute(China-Ukraine E.O. Paton Institute of Welding), Guangdong Provincial Key Laboratory of

Advanced Welding Technology, Guangzhou, Guangdong, 510651, China

Keywords: Robot, friction stir welding, secondary developed, constant force control, control system.

Abstract: In this paper, the composition, principle and method of robot force control for friction stir welding were

studied. Based on the RSI interface of robot and C++ language, a host computer and control system for

robotic friction stir welding was developed, and on-line monitoring of data, displaying of historical data and

preserving of welding- process data were realized. According to the technological characteristics of robotic

friction stir welding, the constant force control of welding-process was realized by combining traditional

PID control and adaptive control. The results indicated that the secondary developed host computer and

control system had the characteristics of simple operation, comprehensive data monitoring, high precision of

force control and stable welding process.

1 INTRODUCTION

Friction stir welding (FSW) is a solid state joining

technology invented by The Welding Institute in

1991 (LONGHURST, W. R, et.al, 2010). Compared

with traditional fusion welding technology, welding

defects, such as crack and void can be effectively

avoided by FSW. And it has the advantages of high

mechanical properties, small deformation, low

pollution and low energy consumption. Therefore,

FSW is especially suitable for joining aluminum

alloys whose melting points relatively low. For FSW,

welding force is a key parameter. And the

indentation, rotating speed and forward speed of the

welding tool will have a significant effects on the

welding force (LONGHURST, W, et.al, 2011;

MELENDEZ, M, et.al, 2003; MISHRA, R, et.al, 2005;

YUSSOF, H, et.al, 2015; ZHAO, X, et.al, 2008). Usually,

larger indentation can induce, larger the contact area

between welding tool and the base material, as a

result, the welding force is larger.

Robotic welding system integrates advanced

manufacturing technologies such as precision,

flexibility, intellectualization, software application

and development. This is especially suitable for

welding work-piece with complex three-dimensional

surface structure. The welding of complex three-

dimensional surface structure is an important-

direction for the future development of FSW. In

addition, constant Z-axial force is a key parameter

for obtaining stable welding quality. However, due

to the limited load capacity and flexibility of the

robot, it is difficult to ensure the steady downward

force of the welding tool on the work-piece in FSW

process. Therefore, it is of great significance to

ensure the constant of Z-axial force for improving

the quality of robotic FSW.

2 COMPOSITION OF Z-AXIAL

FORCE CONTROL SYSTEM

2.1 Composition of Control System

As shown in Figure 1, FSW control system is

composed of a six-axis robot and its control system,

host computer control system, data acquisition and

processing system, force sensor system, spindle

motor, stirring needle, frequency conversion control

system and Bus control system. Six-axis robot and

its control system have high flexibility, high

trajectory accuracy, and thus the processing of

arbitrary trajectory in space can be realized. It is

important for realizing FSW of complex three-

dimensional surface structure. The host computer

control system is composed of operating interface

system and C++ program. The operating interface

444

Xin, Y., Huang, Y., Gao, S., Zhao, Y. and Li, S.

Research and Application of Robot Force Control for Friction Stir Welding.

DOI: 10.5220/0008870104440449

In Proceedings of 5th International Conference on Vehicle, Mechanical and Electrical Engineering (ICVMEE 2019), pages 444-449

ISBN: 978-989-758-412-1

system is used to realize functions, such as welding

process parameters setting, data calling, welding

mode selection, system alarm display, real-time

welding data curve display and data storage and

conversion. C++ program system is used to

communicate with slave system and realize data

exchange. The constant force control of welding

process was realized by the feedback force of C++

program system collected. Data acquisition and

processing system is used to collect and process

force sensing values and transmit to PC by OPC.

Force sensor system are transformed by converter to

measure the Z-axial force value in welding process.

Spindle motor is controlled by frequency conversion

system to realize speed control in welding process.

Bus control system includes Profinet communication

between robot and data acquisition system, RSI

communication between robot and host computer,

OPC communication between data acquisition

system and host computer. It realizes data exchange

between robot control system, host computer system,

data acquisition system and salve system (force

sensor, frequency converter, etc.). These data are

important guarantee to realize the coordinated

operation of the whole system.

2.2 Data Acquisition and Processing

System

Accurate acquisition and processing of Z-axial force

is an important foundation and guarantee for

realizing constant force in FSW process. Data

acquisition and processing system includes force

acquisition (induction) module, signal amplifier,

signal converter module, and AI data acquisition

module and data processing software. As shown in

Figure 2, the force induction module is used as a

sensitive element (mv level induction), which is

amplified by a signal amplifier. And then it is

converted to the values that can be read by the AI

analog module by signal converting module. Finally,

the collected values are processed and converted by

data software to the actual force values. The data

acquisition and processing system regulates sensitive

signals, including functions of impedance matching,

limiting, zero adjustment correction, isolation, and

wave filter (low pass, power frequency and GSM

filter). And finally converts them into 4-20 mA

signals with strong interference. These signals (4-20

mA signals)are collected and processed by

industrial-level PLC analog module. The influence

of gravity changes of the self-weight of spindle

motor and cable that caused by welding angle

change in the welding process are eliminated. As a

result, the accuracy and stability of data transmission

can be ensured.

Frequency Conversion

Control System

Host PC system

Spindle motor

Robotic control system

Six-axis Robot

OPC

RSI

Welding tool

Welded Parts

Force sensor

Data acquisition and

processing system

Figure 1. Composition of control system.

Research and Application of Robot Force Control for Friction Stir Welding

445

OPC

signal

amplification

Force sensor

Signal

conversion

AI module

Host PC system

（monitor and display）

data

processing

Figure 2. Data Acquisition and Processing.

3 CONTROL MEHTODS AND

PRINCIPLES

3.1 Control Principle of Host PC

The host computer is an important human-computer

interface, and it realizes the functions of parameter

setting, data calling, mode selection, data monitoring,

alarm display and historical data reproduction.

Mature interface development software, such as

Kingview, Intouch and Labview, has the

characteristics of mature, simple development and

short process cycle. However, they cannot directly

develop robot program further. Therefore, these

kinds of software can not directly control the

program of the robot and Z-axis force of the robot.

Based on C++ powerful development language,

robot program will be redeveloped by the KUKA

RSI interface. The host computer exchanges data

with Siemens PLC through OPC that is simple in

communication, strong in reliability and high in

stability, so the force value reading can be realized.

At the same time, the host computer exchanges data

between the robot and the host computer through

KUKA RSI interface. The robot program is

redeveloped using C++ language, and its functions

includes reading the position information of the

robot, PID control and other logical control. Closed-

loop of Z-axis force control can be realized by the

force value that is used as the input and feedback of

the parameters.

3.2 Control Method of Z-axis Force

3.2.1 Control Principle of Z-axis Force

Figure 3 is schematic diagram of constant force

control. The communication between the robot and

the host computer is realized through the KUKA

RSI interface, and the robot program is redeveloped

using C++ language. In FSW process, after

comparing setting value with actual value of Z-axis

force, the redeveloped program tunes and adjusts

PID parameters by control algorithm combined with

conventional PID and adaptive control, and controls

and adjusts robotic attitude in real time. At the same

time, considering the timeliness of RSI

communication and force feedback, as well as time-

delay of robot execution, it is necessary to fine-tune

and delay the PID parameters according to the

position parameter information of robot feedback, so

that the difference between actual value and setting

value of Z-axis force can be reduced continuously

and Z-axis force can be kept constant within the

range of +10% to achieve the purpose of constant

force control.

3.2.2 Adaptive Control Algorithm

Traditional PID control is one of the earliest

developed control strategies. It is used widely in

process control and motion control because of its

simple algorithm, strong robustness and high

reliability. According to statistics, PID control

methods account for more than 90% in industrial

control. Traditional PID is also the most used and

mature control method in force control. Common

application situation of force control includes robot

grinding, robot polishing, hydraulic pressure control

test bench and force control of robotic FSW. In

process of force control of robotic FSW, the welding

tool needs to overcome the larger axial resistance

because of its direct contact with welded materials.

At the same time the value of axial resistance is

significantly affected by the amount of force,

rotating speed and forward speed of the welding

tool. So there are many uncertain factors which can

influence the Z-axis force. It is difficult to ensure the

system can obtain well welding performance under

all working conditions by using conventional PID,

even the phenomenon of instability and welding

penetration be occur due to the influence of

inappropriate parameters.

ICVMEE 2019 - 5th International Conference on Vehicle, Mechanical and Electrical Engineering

446

Force setting

C++language

Control Program

Robotic

Program

Secondary

developed

Proportion

Integration

Differential

Executive

Body

Robotic

parameters

Actual force

Force feedback

Position

feedback

Adaptive

algorithm

Parameters

tuning

Force

feedback

Theoretical value

+ -

Figure 3. Control principle of Z-axis force.

In view of the shortcomings of the conventional

PID control methods, and according to the

characteristics of FSW process, a method combining

conventional PID control algorithm with adaptive

control is adopted. This proposed method not only

has the characteristics of high precision, robustness

and reliability of the conventional PID control, but

also increases the flexibility and adaptability of the

system.

The tuning process of parameters of PID

algorithm that adding adaptive control algorithm

mainly refers to the process of pre-setting

parameters and continuous optimization of pre-

setting parameters. Firstly, the pre-setting PID

parameters KP1 (Proportion), KI1 (Integration) and

KD1 (Differential) are determined according to the

setting force. Then, in the process of robot welding,

according to the difference e1 between force

feedback and predicted value (Figure 3), the

optimized PID parameters K'P, K'I and K'D are

determined by comparing, reasoning and choosing.

The optimized PID parameters of this process is

calculated and defined from the parameters that

computed by e1 and summarized by experience.

Finally, according to the e2 that is difference

between actual position and the ideal position of

robot (Figure 3), the final parameters KP, KI, KD

are defined by changing the coefficients of the

parameters K'P, K'I and K'D. In this process, the

increase or decrease of the coefficients are mainly

based on e2 (including positive and negative). The

arithmetic is shown in Figure 4.

4 WELDING PROCESS

RESEARCH AND

APPLICATION RESULT

ANALYSIS

In FSW, the stress of the welding tool is large, this

causes structural deformation of welding equipment

easily and deviation of welding position. Therefore,

it is necessary to control the welding process force

of Z-axial. The control mode of FSW mainly

includes constant displacement and constant force

control. The main purpose of constant displacement

is to obtain a reasonable displacement. However,

due to the sensitive force change in welding process,

a small change may cause a greater force change.

This mode is only applicable to special welding

equipment with larger stiffness. Constant force

control mode can ensure force stably in welding

process, especially for the lack of robotic rigidity,

the constant force control mode can effectively solve

the impact of robotic deformation on welding quality

(YAVUZ, H, 2004).

Comparison, Reasoning

and Selection

Figure 4. Adaptive control algorithm.

Research and Application of Robot Force Control for Friction Stir Welding

447

The two modes of constant displacement (welding

depth was 4 mm) and constant force (force was

3800N) were used in FSW of 6063-T4 aluminum

alloy. Weld force cures are shown in Figure 5 and

Figure 6, and weld formations are shown in Figure 7

and Figure 8. For the constant displacement control,

there are obvious groove on the weld surface due to

the large range of force variation, serious shaking in

the plunge and poor stationarity in welding process

caused by the robotic deformation. And for the

constant force control, the weld performance can be

improved significantly, at the same time, it has some

advantages, such as stationary plunge, high precision

and smooth welding process.

Figure 5. Curve of constant displacement.

Figure 6. Curve of constant force.

Figure 7. Weld formation of constant displacement.

ICVMEE 2019 - 5th International Conference on Vehicle, Mechanical and Electrical Engineering

448

Figure 8. Weld formation of constant force.

5 CONCLUSION

(1)A host PC system that was suitable for robotic

FSW by robotic RSI interface and C++ language

was developed, and the data communication among

the robotic control system, host PC system and PLC

control system was realized.

(2)Considering time-delay of robot execution

and technological characteristics of FSW ,the

redeveloped robotic program combines traditional

PID control with adaptive control, controls Z-axial

force can be kept constant within the range of +7.5%

in welding process, and enhances the welding

quality significantly.

(3)The application results show that the

secondary developed host PC system and program

control system have characteristics of friendly

operability and versatility, comprehensive data

monitoring, and reduce the workload of technicians

greatly.

(4)Through data monitoring, curve display,

historical data calling and welding process data

preservation, this research provides important

methods and data support for the optimization of

welding process parameters and the improvement of

welding quality.

ACKNOWLEDGMENTS

The authors are grateful to be supported by the

National Key Research and Development Program

of China (2018YFB1306404), by the Key areas

Research and Development Program of Guangdong

Province (2015B090922011), by the GDAS' Project

of Performance Appraisal (2019GDASYL-0302013),

and the Science and Technology Plan Project of

Guangzhou City (201704030038, 201704030068,

201807010068).

REFERENCES

LONGHURST, W. R., STRAUSS, A. M., COOK, G. E.

(2010) Torque control of friction stir welding for

manufacturing and automation [J]. The International

Journal of Advanced Manufacturing Technology.

LONGHURST, W. R., STRAUSS, A. M., COOK, G. E.

(2011) the identification of the key enablers for force

control of robotic friction stir welding [J]. Journal of

Manufacturing Science & Engineering.

MELENDEZ, M., TANG, W., SCHMIDT, C. (2003) Tool

forces developed during friction stir welding [J].

NASA Technical Reports Server.

MISHRA, R. S., MA, Z. Y. (2005) Friction stir welding

and processing [J]. Materials Science & Engineering.

YUSSOF, H., MISKON, MF, BUSU, N. (2015). A review

of force control techniques in friction stir process [J].

Procedia Computer Science.

YAVUZ, H. (2004) Function-oriented design of a friction

stir welding robot [J]. Journal of Intelligent

Manufacturing.

ZHAO, X., KALYA, P., LANDERS, R. G. (2008) Design

and implementation of nonlinear force controllers for

friction stir welding processes [J]. Journal of

Manufacturing Science & Engineering.

Research and Application of Robot Force Control for Friction Stir Welding

449