Investigate Weld Quality of Gas Metal Arc Welding Process Through Online Monitor

Nishadi WEERASINGHE, Banuka PANDITHASEKARA, Dulaj THENNAKOON

(Department of Mechanical Engineering, Sri Lanka Institute of Information Technology)

Abstract: Gas metal arc welding (GMAW) is also referred as the metal inert gas (MIG) welding which is a process of welding done by the formation of an electric arc between the consumable wire electrode and the workpiece. Through the welding process, a continuous flow of inert gas is supplied, and it avoids the weld being subjected to react with atmospheric air. The process can be automatic or semi-automatic where the main input parameters like current and the voltage can be direct and constant, respectively. Not only the current and voltage the welding quality depends on some more input parameters such as arc gap, velocity, and temperature.In this paper, we explain about a setup which is capable of real-time monitoring of input parameters mentioned above and selecting the best MIG welding parameters for the mild steel. The setup is composed of several sensors and microcontrollers for the collection and the measurement of the input parameters. The samples were categorized according to the federate and the voltage adjustment of the selected welding machine. Then the final objective was to identify the samples of the weld with different parameter changes which are monitored through the system. For the analysis, the samples were subjected to tensile and hardness tests, and microstructure tests to find the dependence of the input parameters which effect for the weld quality. Finally, the experimental results verified the effectiveness of the system for the selection of the quality weld.

Keywords: Real-time, Metal Inert Gas (MIG), Gas Metal Arc Welding (GMAW), Input Parameters, Monitoring Setup, Weld Quality

1 Introduction

The gas metal arc welding (GMAW) process is commonly used in industrial level as it is highly productive and cost effective. GMAW is the primary welding process used in various applications and it is known as Metal inert gas welding or metal active gas welding process[1]. An electrics arc is ignited between the workpiece and the consumable wire electrode is used to heat and melt the wire. Metal droplets will fall on to the workpiece and form the metal join. Most of the time the continuous GMAW is monitored real time in industrial levels. Due to this, there is a greater possibility to happen different errors in the weld surface which may cause to the quality of the weld. But real-time monitoring of the input parameters of the weld may be a vast support to have further analysis to the weld quality and it will be the best method to select the best range of input parameters to have a better-quality weld[2].

According to the present days advancement in manufacturing filed, online quality prediction is one of the most important and demanded activity. Real-time monitoring of the input parameters of the weld maybe a vast support to have further analysis to the weld quality and also it will be the best method to select the best range of input parameters to have a better-quality weld[12].Mayer[5]. Rostek used computer-aided acoustic pattern recognition to test the monitoring capabilities of acoustic signals[6]. Also observed the effect of operating parameters, such as voltage supply, the feeding rate of consumable wire electrodes, a standoff of contact pipe and flow rate of shielding gas on an arithmetic average six of all frequencies amplitude and squared center of noise spectra gravity. Some authors believe that surrounding noise, which is not a result of the welding process, obstructs analysis of measured signals and might be considered as one of the most significant obstacles for the acoustic monitoring technique[4-6-9]. According to C.S Wu, T. Polte and D.Rehfeldt research study[10], use of a statistical method to analyze the changes of the weld with the necessary input parameters Voltage and Current. E. Cayo and S.Alfaro[11]have proposed a method of welding quality monitoring by the use of acoustic sensing. It is an inspection method which evaluates the Gas Metal Arc welding process using a non-instructive process sensing. Z. Wang[13]established a paper which aims at developing a system which online monitors the GMAW using image processing technique. The proposed image processing method forms the foundation for the measurement of the weld pool shape with closed-form solutions.

This review of relevant aspects of the literature points out the significance of the online monitoring methods and also the quality management of GMAW through input parameters. It addresses different concepts associated with the online monitoring systems.This review identifies the methods which were completed in the past and according to our methodology, it proves that still there is no any completed research through Arduino online quality monitoring setup for GMAW.

Though welding is one of the most common manufacturing processes, the quality of welding is the most considered factor. The quality of a weld may change with the basic requirements of welding such as current, voltage, welding run speed, the distance between the arc and the workpiece and the bonding of the two materials. Welding quality is depended upon the experience of the welder. Quality of weld is depended upon input parameters. It is benefited if there is a system to monitor these input parameters during the welding. Since the welder doesn’t have any definite clue about how the material has bonded. But, if the parameters of the weld are monitored and displayed then the welder will be able to adjust in such a way so that the weld quality will be improved. It is obvious that key factors for the control of the weld pool are the measurement and sensing of the weld pool features.Welding process can be automated with the interface of reliable sensors and the feedback controllers.

2 Theoretical Analysis

2.1 Major Concepts Used

The output of the project is to the analysis of the quality of the GMA weld samples by monitoring the necessary input parameters of welding through the results obtained by an online monitoring system. The research was done by keeping some parameters at constant levels during the welding process which may not possible to measure at each level. That is to obtain correct online measurements through the online monitoring setup for all the testing mild steel samples.

2.1.1 Constant Parameters

i. Welding environment – room temperature

ii. Sample dimensions

● Width -2.5 cm

● Length – for hardness test (2x4cm), for microstructure test(2x2.5cm), for tensile test(2x10cm)

● Thickness – 5mm

iii. Feed wire diameter – 0.8 mm

iv. Inert gas flow rate – 5 l/min

v. Bevel angle – 45 degrees

vi. Welding position – 1G

2.1.2 Measuring Parameters

i. Voltage – 0 to 24V

ii. Current – 40A to 140A (depend on the electrode characteristics)

iii. Arc gap – vary with the welder gun position(2mm-4mm)

iv. Welding speed - depend on the specimen welding distance and the time taken to weld

3 Methodology

The aim of the study is to build up a setup which measures the value of relevant input parameters simultaneously with the welding time. The beginning of the study was started with the study of the welding plant. That is, according to the input parameters which is going to be measured, those parameter ranges should be known. When considering the MIG welding process the most critical two parameters are voltage and the current. But these two parameters vary according to the characteristics of the specimen which is going to be weld and also with the characteristics of the electrode.

Fig.1 Welding Machine

Telwin Masteringmig 270/2type of welding plant shown in Fig.1was used for the study. The specification of the plant as follows, main voltage – 230/400 V,current range – 28 – 270 A, voltage adjustment positions – 10, wire diameter steel type – 0.6 – 1.2 mm.

According to research, the wire diameter of the electrode was chosen as 0.8 mm and the relevant ranges of current were obtained as 40A – 140 A which depends on the electrode dimension and the specimen characteristics which is to be welded. According to the obtained range of current values, the set of voltage adjustments (second and sixth positions) were selected.The maximum voltage of the welding plant is 24V.Therefore, the main task is to select the proper sensors which are capable of measuring the range of voltage levels and also the current levels during the welding.According to the practical sessions, three feed rates as 3, 4, 5m/min at different current positions were selected. The importance of the welding current is to maintain a proper stable arc and also the variability of the welding current primarily controls the amount of the weld metal that is deposited during the welding.Welding voltage controls the arc length which is the distance between the molten weld pool and the wire filler metal at the point of melting within the arc.

The measuring of the arc gap is a difficult task as it is with small values of milli meters. Also, the measurement of the welding speed also depends upon the welder and the specimen’s distance to be weld where it is a function of the time and the distance traveled. The distance traveled is the actual distance where the weld metal is deposited from the starting of the arc and to the ending of the arc. But the welding speed will change when the welding variables such as position, filler metal diameter, joint accessibility changed. Therefore,it is more important to select the proper sensors to obtain the correct measurements before starting the setup building. It is essential to have a clear idea of each sensor function.

4 Sensor Fabrication

Selecting the proper sensors to obtain the correct measurements, was done prior to setup the experimental platform. The setup control is performed by interfacing the sensors to an Arduino microcontroller.Finally, the arranged setup should communicate with the welding process through an Arduino coding and also should be a convenient task to monitor the parameters using the prepared arrangements. The setup arrangement is completed using the following equipment’s which was selected by after handling the number of setup trials.

Two Arduino Uno REV3 Microcontrollers, Motion Tracking Device-(MPU-6050), ACS758 Arduino Current Sensor, Arduino 25V Voltage Sensor Module,VL53L0X Time-of-Flight Distance Sensor, HC-12 Long Range Wireless Communication Module, Arduino SD Card Module, two 3D printed setup covering boxes, 9V Adapter, jumper wires and USB cable serial communication wires, bread boards, Omega Brand 2 Color Fiber Optic Infrared Ratio Temperature Measurement and Control System were used for the set up the experimental platform.

The online monitoring setup consists of two sections where the measurements of the welding speed and the arc gap are taken at the welding gun itself while the voltage and the current are measured at the plant itself. Therefore, it needs to have two Arduino microcontrollers and also needs to communicate within the two setups to monitor the results. To complete the task,using of a communication model was obtained.

4.1 Motion Tracking Device-(MPU-6050) –Welding speed/ Travel Speed Measurement

MPU-6050 shown in Fig.2 is a6-axis motion tracking device where 3 axis Gyroscope and 3 axes accelerometer on the same silicon die. Also, a Digital Motion Processor (DMP)which process complex 6 calculations onboard to minimize errors in each sensor was used. Itis an advantage in Arduino, which eliminates the need to perform complex resource intensive calculations. Through an I2C sensor bus directly accept inputs from an external3-axis compass and also complete 9 axis Motion Fusion output. MPU-6050 support only the I2C serial interface only, not the SPI interface where communication with the Arduino also done through the I2C interface. Once the wiring is finished when the sensor is using for the first time it is needed to install the relevant Arduino library which support the interfacing of the sensor. Therefore, once the calibration was completed a separated program was written as to obtain the velocity from the acceleration data in all three axes.

Fig.2 Motion Tracking Device

4.2 ACS758 Arduino Current Sensor

ACS758 Arduino Current Sensor shown in Fig.3 was used for measuring welding current. This device is consisting of a precision, low-offset linear hall circuit with a copper conduction path located near the die.When the applied current flows through this copper conduction path it induces a magnetic field which the hall IC convert into a proportional voltage. Through a proximity of the magnetic signal to the Hall transducer,the device accuracy is maintained.

Fig.3 ACS758 Arduino Current Sensor

The results obtained by connecting the sensor to a voltage supplier. The accuracy of the sensor, readings was rechecked by installing the sensor to the welding plant itself in series to the current flow and checked for the readings by comparing with the readings obtained with the caliper.

4.3 VL53L0X Time-of-flight Distance Sensor

The Fig.4 shows the device which is capable of measuring accurate distances up to 2m also it is a tiny package with excellent dimensions. It uses the technology of laser emissions to find out the correct distance. It is 1D gesture recognition, and the distance measurements can be read through a digital I2C interface. The time-of-flight measurement enables the sensor to determine the accurate absolute distance to a target without the object’s reflectance significantly influencing the measurements.

Fig.4 VL53L0X Time-of-flight distance sensor

Inbuilt Arduino library for the sensor was used in programming. Sensor calibration results obtained by measuring the distance of 40mm or 4cm.

4.4 HC-12 Long Range Wireless Communication Module.

Wireless Communication Module is a device which is used to build up a communication between the two Arduino boards because the position of the two boards to be in two places. But at the final monitoring,the results or the outputs of both Arduino boards should display at one monitor. Therefore, the selection of this HC-12 module was done because of its higher accuracy in transmission and also the speed of the transmission of data. It is an RF device which can transmit data nearly a half Km in less-than-ideal circumstance. Consists of an antenna, so to have a better transmission it needs to place both the antennas of the receiver and the transmitter faraway from each other.

4.5 Arduino SD Card Module

Arduino SD Card Module is a device can be used as a simple solution for transferring data to and from the SD card. The pinouts directly compatible with Arduino but can be used in other microcontrollers too.This module has SPI interface which is compatible with any SD card and it uses 5V or 3.3Vpower supply which is compatible with Arduino UNO/Mega. This device is used to obtain the online monitored data to an excel file to use for further analysis regarding the quality of the weld.

Box 1(Fig.5), fixed at the welding gun includes the MPU 6050 sensor- measuring welding speed,VL53L0X sensor – measuring arc gap, Arduino Uno board, HC-12 transmitting module, 9V battery, bread board strip and jumper wires. The schematic wiring diagram of the components is shown in Fig.6.

Fig.5 3D Printed Box 1 with the Components

Consider the setup arrangement at the welding gun, the included sensors are MPU-6050 which is used to measure the travel or the welding speed. Since the sensor is measuring the acceleration in all 3 axes as x, y,and z, it needs to obtain the velocity through the acceleration values. Then according to the sensor characteristics, the acceleration in each direction is equal to a value (equation 1) which is the velocity component in each direction that is given by equation 2.

The final velocity is calculated by taking the square root of each direction square velocity.

Fig.6 Wiring Diagram of the Components Found in Box 1

The VL53L0X sensor is used to calculate the arc gap, which is about 2 to 4 mm. The calculation of the final arc gap is done manually by considering the obtained results for the minimum distance. Arc gap is considered as the distance between the electrode and the workpiece which is shown in Fig.7. Keeping the welding torch distance constant while only varying the electrode tip geometry leads to considerable different arc length. It concludes that reducing the electrode length, leading to increasing the arc length. The arc length is calculated manually after obtaining the total length and subtracting the minimum distance from each set of distances obtained within the period of the weld which gives the result as the arc gap.

Fig.7 Arc Length Variation in Dependence on Electrode Geometry at Constant Welding Torch Distance[17]

Box 2 - Fixed near to the welding machine as shown in Fig.8, includes ACS758 Arduino Current Sensor, Arduino 25V Voltage Sensor Module, Arduino Uno board, HC-12 receiver module, SD card logger, Bread board strip and jumpers. The schematic wiring diagram of the components is shown in Fig.9.

For the current sensor assembly, it is needed to include the sensor in series connection to the current flowing path as to obtain the welding current. And the sensor is connected in-between the plates which are detached as shown in Fig.10, using two high gauge wires with lug joints. Therefore, the current flowing path is through the current sensor which is needed to connect to the forward and the reserve path of the sensor.

Fig.11 Schematic Diagram of the Online Monitoring Setup

5 Results and Analysis

This section discusses the results obtained from the online monitoring system which show the measured values of input parameters like, voltage, welding speed, current and arc gap. Also, the results obtained by the lab testing of the welded samples welded during online monitoring.

The first set of results are the calibration results of the sensors before taking the final results of the online monitoring.

Calibration plot of Voltage sensor shown in Fig.12 using the method of one point calibration. The results were obtained by measuring the voltage of 9V battery.

Fig.12 Voltage Sensor Calibration Plot

Mpu-6050 sensor- The stability of the results will attain after few seconds and initially, some sort of calibration algorithm is run by DMP as shown in Fig.13.

According to calibration results of VL53L0X, the sensor accuracy is measured and the observations shows the sensor accuracy is at a considerable level where most of its coordinates lie on the best fit line as shown in Fig.14. The accuracy of the sensor is almost+-2 mm. It uses laser technology and gives the least count s 1mm.

Table 1 shows the calibration values and the offset measurements to figure out the sensor accuracy of VL53L0X sensor.

Before welding as the first step the categorization of the samples for welding was done. Consider about three feed rates as 3m/min, 4m/min and 5m/min. Adjustment of voltage positions as, 2nd position, 3rd position, 4th position, 5th position, and 6th position.Experiment types to check the material properties of the welded samples are, Tensile test, Microstructure test and Hardness test.

Initially the samples were categorized accordingly to conduct the material testing lab practical. the total number of samples were 15 pairs foreach test, that is 45 pairs for all three tests. These tests were caried out to obtain the best welding samples where it is with the best results from all three material tests.

Fig.13 Plotted Data for Calibration of MPU-6050

Table 1 VL53L0X Sensor Calibration Results

Fig.14 VL53L0X Sensor Calibration Plot

Through the above lab experiments, it is easy to figure out the material property changes with the welding process. If the sample is having best features,that is, during the welding process the harm to the mother material of the sample is at minimum level.

Accordingly, the input parameters of GMAW process are stored and listed for each and every weld sample. Some of the obtained results are mentioned Table 2 for the first sample set of hardness test - Voltage position -2, Feed rate -3.

The arc gap is calculated by subtracting the distances from the smallest value from the monitored measurement. The velocity is calculated, by considering the time taken to weld and the distance of the weld taken place. Then according to the calculated velocity as listed in Table 2 selected the factor to multiply the obtained velocities from the sensor output.

For Tensile experiment used the Universal testing machine and for prepared the work pieces so that it would give the clear results. Prepared a sample of the mother material (mild steel) with the dimensions of 300 x 23 x 5 mm and for the welding prepared two samples of 150 x 23 x 5 mm. The set of the testing samples were arranged accordingly as shown in Fig.16.

Table 2 Online Monitored and Calculated Results of the Hardness Test Sample

Fig.15 (a) Variation of Current with Voltage (b)Arc Gap with Velocity in 2_3 Hardness Test Sample

Fig.16 Welded Samples for Tensile Test

Table 3 Summarization of the Obtained Tensile Test Results

The tensile test was carried out to find the maximum yield strength of the sample, which would ultimately help in determining the better parameters for welding. Specimen 2, 3 and 6 were failed welding where the metal is not bounded. During the testing procedure, specimens 1, 4, 5 and 9 were not identified as valid welds by the machine since the reading values were comparatively very small. The tensile load that the MIG welding was able to withstand was less than that of the mother material. (Mother material could withstand 675MPa without attaining the breaking point while the maximum tensile stress that a weld could withstand was 485MPa). Specimen10, 14 and 15 can be identified as successful welding’s through tensile test according to the tensile test results as in Table 3 and to arrive at final conclusion it is need to compare the readings with microstructure and the Hardness test also. The ductility of the welded specimen was less compared to the mother material. Apart from the formation of the weld, the specimen was subjected to high temperatures during the welding process. This could have changed the chemical properties of the material which resulted in making the material less ductile.

The set of samples prepared for the hardness test is shown in Fig.17. Similar to the tensile strength in the weld is heterogeneous due to the temperature gradients and the chemical gradients that evolve during the process.

Table 6 Microstructure Results for Selected Samples

According to both tensile and Hardness test results the best characteristics of good quality weld is given by samples 5/3, 6/4 and 6/5.

Fig.23 Online Monitored Measurements of 6_4 Sample for (a) Hardness (b) Tensile (c) Microstructure Test

Dataset of online monitored input parameters for the 6_4 sample welding for all three tests is shown in Fig.23. Also, when considering microstructure of good quality, weld shows in sample 6_4 as shown in Fig.24 and Fig.25. So, the final conclusion is, the best set of input parameters to (5mm thickness & 25mm width plate) mild steel gives by sample 6_4 when considering mechanical properties of the weld. The Table 7 list down the multiplying factor for the calculation of velocity and the arc gap of all three 6_4 samples for the 3 experiments.

Table 7 Date for Calculation of Velocity and Arc Gap for 6_4 Sample

Fig.24 Microstructure View of 6_4 Sample- heat Affected Area

Fig.25 Microstructure View of 6_4 Sample-parent Material

Microstructure of weld is very fine perlite, the Tensile stress of sample 485MPa (UTS), Hardness value 115(HRC), current range 70-100 (Amps), voltage range 23-34and the arc gap range 2-5 (mm).

6 Conclusion

With this project study it is concluded that the online monitoring setup arrangement is a complex task and fine task to be completed which is mainly connected with the input parameters of the weld. The results obtained from the online monitoring system was compared with the results measured manually and the quality of the weld was analyzed with the obtained test results by the lab testing and compared the features of the weld with the variation of the input parameters. The research found the GMAW input parameter measurement through the online monitoring systems can be considered as a convenient method to work in the analysis of the quality of the weld. During the monitoring, the variation of the readings is at higher levels,therefore it gives out all parameter variations accurately. The higher accuracy of the sensor measurements shows the variation of the current and voltage at low and high levels.

For the future work, the setup can be recommended to develop with wireless communication if a Wi-Fi module is interfaced with the setup.