Need a quote for a coil or heat exchanger design or service?
Request Quote

Analysis of the Effects of Heat Input and Weld Bead Size On Welding Distortion For GTAW and GMAW-P

May 10th, 2022
 PROJECT NAME

Analysis of the Effects of Heat Input and Weld Bead Size On Welding Distortion For GTAW and GMAW-P

DEGREE

Master of Science in Welding Engineering (Non-Thesis)

COUPON WELDING

Tony Meyers – Welder

WRITTEN BY

Christian Hallingstad

COURSE NO.

WELDENG 7193.02

TECHNICAL ADVISEMENT

John Edwards P.E. – Engineering

Pat Lucey – Plant Foreman

ADVISOR

Dr. David Phillips

NO. OF

PAGES 16

PROJECT SPONSOR(S)

Jason Thomas – IHT President

Paul Nelson – IHT Vice President

ADDITIONAL COMMITTEE MEMBER Dr. Avi Benatar CREDIT HOURS

3

 

 

REVISIONS
Rev DESCRIPTION BY DATE APPROVED BY
Initial document creation and release. CJH 12/27/2021
A Added titles to figures, added details about weld coupon measurement techniques and dimensions, changed report title, updated graphs and charts for clarity. Labor savings example added to Page 13. CJH 1/3/2022
B Added initial values for x-variables on Page 4 CJH 1/5/2022

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 1


 

 

TABLE OF CONTENTS
ABSTRACT
3
INTRODUCTION
3
METHOD OF EXPERIMENT
4
RESULTS
6
GMAW-P
6
GTAW
9
GTAW VS. GMAW-P
10
CONCLUSIONS
12
CONTINUED PROJECT WORK
13
REFERENCES
16

 

 

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 2


 

ABSTRACT

Industrial Heat Transfer, Inc. (IHT) has been manufacturing fabricated pressure vessels for various industries for the past three decades. Since IHT has been performing welding for an extended period, it has a lot of experience with earlier versions of welding processes. With the advancements in welding technologies, new welding techniques and strategies may be employed to improve overall welding process and final weldment design and functionality. The purpose of this paper is to compare the effects of weld bead size (throat of a groove weld) and welding process (both gas metal arc welding with a pulsed spray transfer mode (GMAW-P) versus gas tungsten arc welding (GTAW)) and how they influence weldment distortion.

INTRODUCTION

GTAW is a reliable welding process that allows the welder the most control over their welding variables. This welding process is especially useful for ensuring proper penetration into the root pass of a weld joint. The root pass is arguably the most important weld pass because it is the first line of defense to create a proper water, air, or other media-tight seal, as well as holding pressure on the other side of the weld joint. If this weld pass is not properly placed by the welder, there are several adverse effects which may stem from an improper fusion of the base metals, such as a stress concentration on the pressure-side of the weld joint, which may allow a crack to manifest and propagate through the weld joint, resulting in a catastrophic failure of the pressure vessel. GTAW is also a superior process when space restrictions are present. GTAW allows for several different torch designs that allow the welder varying profiles that allow for tighter

 

welding areas. The variety of GTAW torch designs is beneficial compared to GMAW-P because traditionally, GMAW torches are bulkier and less dynamic.

However, although GTAW allows the welder more control over their weld, it is also a more tedious and slower process, which decreases productivity and increases labor costs. There are several welding variables that all contribute to the complexity of GTAW. In my opinion, this makes the GTAW process the most skill-based of the welding processes. IHT utilizes several GMAW process transfer modes, such as spray or short-circuit, but with the rising popularity of pulsed-spray, and the advancements in synergic inverter welding machines, IHT engineering initially set out to determine the viability of utilizing GMAW-P in areas where GTAW was traditionally employed, or areas where GMAW-Spray could not be used due to heat input or welding position. GMAW-P was initially believed to be a hybrid blend of GTAW-like control, with more freedom with welding positions than GMAW-Spray, less heat input than GMAW-Spray, the penetration profile of GMAW-Spray, and with less weld spatter than GMAW-SC (short circuit).

A few reasons IHT initially added GMAW-P into their welding repertoire was to find a replacement for GMAW-SC on stainless-steel to mitigate, or prevent altogether, weld spatter that required unnecessary grinding and additional labor. Another application was to have the travel speed and productivity of GMAW, but with the intricacy of GTAW when welding a longitudinal seam butt joint on a split pipe header. When the header design gets to be longer than a few feet, GTAW is a very laborious process, which slows down production and decreases throughput.

Traditionally it was believed that GTAW had to, at the very least, be used on the root pass of a weld joint to ensure a leak-tight seal. Then, if

 

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 3


 

desired, other welding processes could be utilized for the subsequent passes. There is also the belief that GTAW produces the least amount of welding distortion compared to other welding processes utilized at IHT. This experiment is not stating that GMAW-P can be employed as a direct replacement to GTAW in all scenarios, but it is comparing the two processes to highlight that GTAW is not the only welding process that can produce a sound root pass or the only process that can mitigate welding distortion. Another reason for pursuing the GMAW-P option is to improve productivity by not only increasing travel speed during arc time, but by minimizing the number of settings and variables for the welder to account for. This will decrease machine setup times and decrease overall welding times to help decrease labor costs and improve overall shop throughput. weld-coupon model
 

METHOD OF EXPERIMENT

This experiment will utilize a groove-butt weld joint welded in the 1G position, which joins two pieces, each 1/2- inch-thick pieces of ASTM A36 steel bar stock. Each piece of steel measures five inches long, which is the weld bead length, and 6 inches wide. A 1/4-inch-thick backup strip was used because the weld was welded from one side only. Both bar stock pieces used in the weld coupon assembly were beveled to a 45- degree angle with a knife edge to promote better fusion and a 1/8-inch root gap. Each coupon assembly was mechanically stamped either with a “G” for GMAW-P or a “T” for GTAW, as well as a numerical designator to track each sample. A representative model of the weld coupon assembly used in this experiment can be seen in Figure 1.

 

 

In this experiment, the overall weld size is not considered because that is determined by the requirements of ASME Section VIII, Division 1 Code, or other applicable specifications, and is therefore disregarded. The individual weld bead size that makes up the entire weld size is a consideration in this experiment and is used as a comparison to see its effect on weldment distortion.

Each weld coupon assembly was marked with a punch at three locations on each side of the weld path. These locations were spaced equally apart in the x-direction (direction of weld path) and the y-direction (perpendicular to weld path). A sketch of the punch marks and their layout can be seen in Figure 2. Initially, the distances between punch marks (A1, A2, and A3) were all equal to y0 which was two inches, and the values for y1, y2, y3, and y4 were all equal to a value of 0.75-inches. Also, x0, x1, and x2 all had equal initial values of 1.50-inches.

 

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 4


All weld coupons were marked the same way and measured after tacking the weld coupon assemblies together, but prior to laying weld passes. These marks will be measured postwelding to determine the magnitude of welding distortion in the longitudinal direction (along weld path) and in the transverse direction (perpendicular to weld path) by measuring the values of “A” and “x” after welding. These final values were subtracted from the original values to obtain the final distortion kilojoules per inch, kJ/in.) were calculated. Equation 1 shows how the travel speed was calculated. The time welding each weld pass was measured using a cellphone stopwatch. Equation 2 shows how the heat input was calculated for each weld pass. This equation was taken from ASME Section IX, Part QW409.1; “P” is power and has units of Kilowatts (kW) (American Society of Mechanical Engineers, 2021).
Similarly, after tacking the weld coupon assemblies together, they were placed on a granite surface table on the bottom surface of the backup strip. The heights at various locations on the top surface of each piece of bar was measured. These measurements were measured again post-welding to determine the magnitude of angular distortion. A Vernier height gauge was used to measure the top edge of the parts, at a location parallel with the line connecting the A3 punch marks, then this value was subtracted from the initial measured value to get the distortion. See Figure 3 for reference.
After measuring and recording those variables, a diagram of each weld pass was recorded to show approximately the size and location of each weld bead. An example of the weld pass diagram can be seen in Figure 4, which is the diagram for a 3/8-inch weld bead for the GMAW-P process
To compare the effects of weld bead size, which determines the number of passes per weld joint, as well as welding process during each pass for each weld coupon assembly, the values for amperage, voltage, and arc time (time welding) were measured and recorded. From these values, the travel speed (S, unit of inches per minute, ipm) and heat input (HI, units of
The welding process machine settings were also recorded for each welding process and each weld bead size. Those settings can be seen in Table 1. Amperage (A) and voltage (V) are not

 

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 5


 

settings on the GMAW-P welding machine. The GMAW-P welding machine used in this experiment was a Miller Millermatic 255. Since this is an inverter-style synergic machine, the only settings the welder set were the arc length and the wire feed speed. The other settings are automatically adjusted during the welding process. was used. The size of filler metal used for each weld bead size and welding process is also listed above in Table 1. The type of filler metal used was ER70S-6, per specification ASME SFA-5.18. Between each subsequent weld pass, the toes of the welds were mechanically ground to ensure proper fusion between the new weld and previously laid weld metal. After grinding, the welds were brushed with a steel wire brush, then acetone cleaned to ensure a nascent surface was prepared to mitigate any foreign object debris (FOD) from contaminating the welds.

The shop area where welding was performed has the ambient temperature maintained at 70°F, so preheat of the steel material was not employed prior to welding these coupons. Specific interpass temperature was not measured, but interpass temperature never exceeded 350°F; this was verified using a temperature crayon. The coupons were also welded in an unconstrained manner, which is perceived to decrease the residual stress in the coupon, but to know for sure this would be measured via the hole drilling method (Group, 2010).

RESULTS

After all welding was completed between the three weld bead sizes for both welding processes, the data was analyzed and plotted in several different graphs.

GMAW-P

One comparison was plotted showing the effect of travel speed on heat input. From Figure 5, all weld bead sizes trended downward, meaning the faster the travel speed, the lower the heat

TABLE 1 NOTES:

1. A and V for GTAW weld sizes are averages of recorded values.

2. Arc Length setting on machine is dimensionless.

The type of shielding gas that was used for GMAW-P was 95% Argon – 5% CO2, with a flow rate of 35 cubic feet per minute (cfm) using an argon flow scale. The type of shielding gas used for GTAW is 100% Argon, with a flow rate of 20 cfm using an argon flow scale. GTAW utilized a 2% thoriated tungsten electrode (EWTh-2) per ASME SFA-5.12 and the current type was direct current straight polarity (DCSP, electrode negative). The gas cup size for GTAW was 1/2- inch diameter. For GTAW, IHT welding procedure specification (WPS) P1GT Rev. 2 was used and for GMAW, IHT WPS P1GMP Rev. 0

1 50 is the factory preset value. By increasing the arc length, the current is increased which increases the melting rate, which requires a higher wire feed speed (Guide to Pulsed MIG Welding In Manufacturing, n.d.).

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 6


 

input. This is in-line with the initial hypothesis, so these results were expected.

The initial hypothesis was that GMAW-P allows for a faster travel speed, which would result in lower heat inputs and thus less welding distortion. Also, IHT would experience greater productivity due to less arc time, which would decrease labor costs.

Table 2 is a comprehensive average of the different types of measured distortions2 for each weld size and sample. As shown, even though travel speed is much faster between 1/8-inch weld beads and 3/8-inch weld beads, for example, which would result in a lower heat input, since there were 10 more weld passes in total for the 1/8-inch weld bead size, this resulted in more distortion of the weld coupons. The most considerable distortion is angular distortion; the longitudinal distortion can be considered negligible for GMAW-P, and the transverse distortion is consistent between weld bead sizes.
Figure 6 shows a graphical comparison of the different types of distortion and how the number of weld passes affects them. This graph shows, on average, the less amount of weld passes on the weld joint, the less amount of distortion that is experienced. There are other variables that should be taken into consideration regarding distortion versus weld passes, but that additional experimentation will be for future projects.

2 Average Angular Distortion is y3 and y4 from Figure 3. Average Longitudinal Distortion is x1 and x2 from

Figure 2. Average Transverse Distortion is A1, A2, and A3 from Figure 2.

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 7


 

Table 2 NOTES:

1. Average Angular Distortion is y3 and y4 from Figure 3. Average Longitudinal Distortion is x1 and x2 from Figure 2. Average Transverse Distortion is A1, A2, and A3 from Figure 2.

Travel speed is another variable that plays a large role in welding distortion and welding productivity. Per Equation 2 from ASME Section IX, if travel speed is increased, the heat input will decrease. Figure 7 shows a comparison of the number of passes and its effect on travel speed. The number of passes is directly related to the weld bead size as well. So, the larger the number of passes, the smaller the individual weld bead size is.

Therefore, by making individual weld beads smaller, less weld filler metal is deposited per pass, and the gun is able to travel faster along the weld path. However, even though travel speed is increased for each weld bead, the quantity of weld beads is also increased, which ends up taking more time in the long run. Figure 8 shows the total arc time based on how many weld passes the sample required. As can be seen, the less weld passes the joint requires, the less total arc time the sample experiences.

Referring to Figure 5, the less amount of weld passes, the higher the heat input tends to be. This is due to the larger amount of heat flux (energy per unit area) required to make a larger weld pool and to melt the filler metal coming out of the gun at a higher feed rate. But since there are fewer weld passes, the sample experienced less overall weld distortion after cooling. Neglecting the time spent cooling between individual weld passes, this may be due to the amount of total arc/energy time the weld coupon is experiencing. Since the coupon is experiencing longer periods of heat flux from

 

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 8


 

more weld passes, that additional heat is having a significant impact on that sample’s distortion. Overall, the data shows that a larger weld bead size, resulting in fewer weld passes, is favorable for welding distortion with a bonus of improving shop productivity.

GTAW

Similarly to how GMAW-P was analyzed, the comparison between the effect of travel speed on heat input can be seen in Figure 9 for GTAW. The test results show all weld bead sizes trended downward (with the exception of some outliers), meaning the faster the travel speed, the lower the heat input. This is in-line with the initial hypothesis and the same trend as was seen across all GMAW-P weld bead sizes.

Table 3 is a comprehensive average of the different types of measured distortions for each weld size and sample. The 1/8-inch weld bead size sample was not performed due to time constraints because of the amount of weld passes it would require and may be considered for future projects. Tests showed even though travel speed is much faster between 1/4-inch weld beads and 3/8-inch weld beads, for example, which would result in a lower heat input, since there were eight more weld passes in total for the 1/4-inch weld bead size, this resulted in more distortion of the weld coupons. The most considerable distortion is angular distortion; the longitudinal distortion and the transverse distortion, although still relatively small compared to angular distortion, were also larger than compared to GMAW-P. The different types of distortion, as related to weld passes, can be seen in Figure 10.

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 9


 

Table 3 NOTES:

1. Average Angular Distortion is y3 and y4 from Figure 3. Average Longitudinal Distortion is x1 and x2 from Figure

2. Average Transverse Distortion is A1, A2, and A3 from Figure 2.

GTAW average travel speeds versus number of weld passes can be seen in Figure 11. Just like with GMAW-P, the fewer number of weld passes caused the weld bead size to be larger and resulted in slower travel speeds.

This is in part due to the efficiency of metal deposition between the two processes. GTAW utilizes a non-consumable electrode, which is 2% thoriated tungsten, that passes the welding arc through it to the workpiece. Then the filler metal is pushed into the arc and indirectly melted by the welding arc. As stated in the Principles of Arc Welding Systems, from WELDENG 7001 lecture notes, “for a gastungsten arc operated DCEN, this loss of heat is of the order of only 10 to 20 percent,” (Richardson & Farson, 2006). Whereas GMAW utilizes a consumable electrode arc because the filler metal passing through the welding lead and gun acts as the electrode. This is much more efficient because the electrode melts and is transferred as it is being consumed, which is why it is a higher productivity process. A visual representation of both electrode types can be seen in Figure 12.
If the travel speeds and resultant heat inputs are compared between GTAW and GMAW-P, for two weld bead sizes (1/4-inch and 3/8-inch), it is clear how much faster of a welding process GMAW-P is and the lower heat inputs it produces. Figure 13 shows this comparison. Figure 14 shows a comparison of the different types of distortions for the 3/8-inch weld bead size, between both GTAW and GMAW-P. The

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 10


 

longitudinal distortion is consistent across both processes and is small enough to be considered negligible. Transverse distortion seems to be slightly higher in GMAW-P than in GTAW, but angular distortion is considerably higher in GTAW than in GMAW-P. Since the numerical value of the transverse distortion is around 1/16-inch, we can focus on the angular distortion due to its larger magnitude. In terms of percentages, the 3/8-inch GMAW-P coupons saw an average of 27% lower angular distortion values, but an average of 45% higher transverse distortion values when compared to the GTAW process. For the 1/4-inch weld bead size, the GMAW-P coupons saw an average of 43% lower angular distortion and 26% lower transverse distortion compared to the GTAW coupons.

For the 1/4-inch and 3/8-inch weld bead sizes for both GTAW and GMAW-P, Figure 15

compared to Figure 14 shows the disparity in the angular distortion that takes place when the amount of weld passes differs. As the difference in number of weld passes increases, the difference in distortion is increased as well. This is why it is hypothesized that 1/8-inch weld beads would result in an even greater difference in weld distortion between GMAW-P and GTAW.

 

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 11


 

Lastly, two weld samples, G5 and T5, were sectioned in half using a band saw, then polished using a 60 grit, followed by an 80 grit, sanding pad, and finally etched using a 2% Nital solution. The results of this etching can be seen in Figure 16: 1/4-Inch GTAW Etch and Figure 17: 1/4-Inch GMAW-P Etch. This etching process is very rudimentary, which only shows the line of fusion in both samples and a few of the individual weld passes in the GMAW-P sample. What can be seen is good penetration on the edges of the beveled plates (weld interface) for both processes, as well as good root penetration for both processes. It appears that the GMAW-P process, in this case, had slightly less penetration of the root pass into the backup strip than the GTAW process did. The higher heat input from GTAW, which is a result of much higher amperage, allowed for a better penetration profile than the GMAW-P. In this case, the power input could be adjusted for the welding process to specify a higher amperage for the root pass(es) of the weld joint, then adjust the parameters for the subsequent cover passes, rather than keeping the same machine settings for all weld passes

 

The lower penetration in the GMAW-P root pass may be due to a fast travel speed without a higher power input. Whereas the GTAW sample has a higher power input with a slower travel speed. Therefore, to get a deeper penetration profile, while achieving a fast travel speed, a higher power input is required for the root pass using the GMAW-P process. This will be similar in energy per unit length of weld as the GTAW sample. That way the heat inputs are still relatively equal, while having a much higher melting efficiency and deposition rate. Another observation from the etched samples is the amount of angular distortion you can visually see between them. The GTAW sample has a much steeper plate angle than the GMAW-P sample does, which agrees with the measured data.

CONCLUSIONS

Both welding processes compared in this experiment have their pros and cons depending on their application. In IHT’s experience, GTAW is employed on intricate weldment designs, or weld joints that are inherently small due to code specification requirements, such as those outlined in MIL-STD-22. GMAW-P is employed when welding longer weld beads or larger weld beads are needed. Sometimes the added peace of mind of performing a sound GTAW root pass is needed on higher profile, critical weld joints. But when the weld joint is excessively long, GMAW-P seems to have enough process control and penetration to perform a sound root pass.

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 12


 

In the direct comparison of performing various weld passes at specified sizes for a flat welding position (1G) of a groove butt weld joint, the data shows GMAW-P being a better option to produce favorable weld distortion and higher productivity rates.

As an example, for welding a longitudinal weld joint on a standard pipe header, utilizing GMAW-P over GTAW will produce approximately 58% higher efficiency, which would equate to approximately $10 per pass of labor savings (in this case). If the weld joint requires three passes, that is a savings of $30 per weld joint, or $60 per pipe header, and $120 for the entire unit, which is roughly equal to having one labor hour in savings per unit. A typical order of these units is for a quantity of eight, which is a savings of almost $1000 for the entire order just by switching one welding process for another for one weld joint.

The etched specimens shown in the preceding figures show penetration profiles and lines of fusion being similar between both welding processes. As stated previously, if improvement of the penetration profile of the root pass of the GMAW-P sample is desired, an increase in the power input proportionally with the travel speed would produce a similar heat input as the GTAW sample. This way, both samples would have the same heat inputs, but the GMAW-P would still be a higher deposition rate and melting efficiency. This would just be for the root pass though, then the subsequent cover passes could be adjusted back to the experimental settings that produced favorable travel speeds and distortion values. A comparison of travel speeds and power inputs can be seen in Figure 18: Penetration Profiles For Differing Power Inputs taken from WELDENG 7001.

In this example, both samples were welded with the same welding process with the same heat inputs, but the weld on the right had a higher power input and faster travel speed compared to the weld on the left. Thus, the penetration profile was deeper, and better fusion with a narrower heat affected zone (HAZ) (Richardson & Farson, 2006).

There are still applications for IHT where GTAW is a superior welding process over GMAW-P, such as intricate welds or welds made in space restricted areas. However, if a weld can be made in the 1G position, and the weld is sufficiently large, GMAW-P is a more favorable welding process due to its melting efficiency, distortion control, and higher productivity.

CONTINUED PROJECT WORK

Due to time constraints, the 1/8-inch weld bead size for GTAW was not performed because it would have taken approximately 30 or more weld passes to complete the joint. It was hypothesized that the welding distortion would have been the largest for this weld bead size, due to the number of passes it would require, by looking at the trends displayed in Figure 10. As an addition to this project, the 1/8-inch weld bead sample could be finished to see what the distortion actually is and how it compares to GMAW-P and the other weld bead sizes made using GTAW. In addition to that, a larger number of samples could be made to get a better average for all variables that were

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 13


 

measured. A larger sample size will make these findings more concrete.

Future comparisons in welding processes could also be made between GMAW in the shortcircuit transfer mode, particularly Miller’s Regulated Metal Deposition (RMD) mode and Lincoln’s Surface Tension Transfer (STT) mode. With further advancements in welding machine technology, the progression of the short-circuit transfer mode has allowed for this welding process to evolve from the entry-level GMAW process it once was. These new versions of GMAW-SC allow the welder to bridge larger root gaps than on older machines, have better penetration and root pass fill compared to previous GMAW-SC machines, while still maintaining quick travel speeds and metal deposition rates. However, for IHT to compare one or both of these new GMAW-SC modes, contact with the welding supplier will need to be made to see if a demonstration machine is available for testing.

Another variable to measure during a future experiment would be to see what the effect of a specific interpass temperature has on weld distortion. Would allowing the plate to cool to ambient temperature be more beneficial compared to either performing consecutive weld passes or maintaining the specified interpass temperature outlined on the WPS? In connection with this idea, a metallurgical analysis could be performed on the steel weld coupons to compare the microstructure of the coupons. There may be differences in microstructure, which influences the phase transformation of the metal, depending on the heat input exhibited on those samples and how much the samples are allowed to cool as well as their cooling rates. Depending on the temperature gradient the material experiences during and after welding, and the rate the material is allowed to cool at, the material may exhibit one microstructure over another, such

as body-centered cubic (BCC) versus facecentered cubic (FCC). Figure 20 is an example of a binary phase diagram that may be used to predict phase transformations. Figure 19 shows a visual representation of the type of solidification mode of the material based on the temperature gradient and solidification rate after welding. Since this material is only 1/2- inch-thick, the thermal gradient between the center of the part and the hotter surface of the part should not be too high. A faster solidification rate, such as quenching, would result in a finer grain structure, which would increase the strength of the material.

As stated in WELDENG 7101, reducing grain size acts as a barrier to dislocation motion, which increases strength and toughness (Lippold, 2019). According to the Hall-Petch equation, Equation 3, the yield strength is a function of grain size, and shows an increasing yield strength (σY) with a decreasing grain size

 𝜎𝑌 = 𝜎0 + 𝑘𝑑 −1/2

Equation 3

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 14


 

This is due to the grain boundaries impeding the motion of the dislocations since the dislocations must change directions when coming upon a grain boundary. Therefore, materials with smaller grain sizes restrict the motion of the dislocations more (Lippold, 2019). This motion restriction improves the material’s yield strength, which in turn reduces the probability of plastic deformation. However, if allowing the weld coupon assemblies to cool completely to room temperature, via natural convection and conduction, would that result in maintaining larger grains, much like an annealing process does? If this is the case, the larger grain size of the slower cooling rate samples could potentially have lower tensile and yield properties than a sample with different cooling rates and interpass temperatures. Determining if there are any changes in the material’s microstructure could provide more insight into the viability of certain welding processes on pressure vessels that go into various service conditions.

Another future development to this project would be to use different sizes of filler metals for GMAW-P depending on the weld bead size

specified, much like what was used in this project for GTAW. Further improvement in the metal deposition rate and heat input could be made for the 3/8-inch weld bead had 0.045- inch filler metal been used instead of 0.035- inch. Using a larger filler metal size for GMAWP, like what was done for GTAW, might be a more equal comparison between the two processes. While the deposition rate might increase, it might take more amperage to melt the larger amount of material being used in the GMAW-P process, which would increase the heat input, and potentially the distortion, but how much those variables change is unknown until an experiment is conducted. Lastly, another experiment could be conducted to see how the assemblies react when they are constrained in a fixture. However, this constraint could induce potentially harmful residual stresses in the material, which may or may not be unfavorable during service conditions. A measurement of those residual stresses could be another addition to this added experiment, measured via the hole-drilling method as shown by Vishay Precision Group. With the information gathered from this experiment, and any information gathered from any future experiments that stem from this project, these findings can be added to IHT’s internal welder workmanship training (WWT) program as well as any IHT standard engineering specifications (SES) that are used in production. This information may include actual data on welding distortion and how the effects of weld bead size, maximum weld bead width, and weld bead placement all play a role in part geometry dimensional conformance postwelding. While the welder is limited to variables outlined in the specified WPS, this information is helpful general knowledge for the welder, so they are better equipped to think critically about their welds, rather than simply joining two parts together.

 

 

 

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 15


REFERENCES

Group, V. P. (2010). Measurement of Residual Stresses by the Hole-Drilling* Strain Gage Method.

Guide to Pulsed MIG Welding In Manufacturing. (n.d.). Retrieved from Miller: https://www.millerwelds.com/resources/article-library/an-introduction-to-pulsed-gmaw

Lippold, J. C. (2019). Basic Metallurgy Principles. WELDENG 7101 Lecture Notes.

Richardson, R., & Farson, D. F. (2006). WELDENG 7001 Lecture Notes. Principles of Arc Welding Systems

 

HALLINGSTAD                                               WELDENG 7193.02                                                              P a g e | 16