Development of Early ERW Seam Processes for Line Pipe

Welding textbooks [e.g., see 1] indicate that although flash welding heats the faying surfaces by combined resistance and arcing it is classified an electric resistance welding (ERW) process, and as such is included in this QR segment considering early resistance welds. Regardless of which among the early autogenous processes were used to make line-pipe, all electric-resistance long-seam butt welds are made by the coupled effects of heat and pressure applied across the seam of the pipe. The coupled effects of heat and pressure produce an upset local to the heat affected zone (HAZ) of the weld that is trimmed (scarf’d off), as elaborated in References 2 and 3.

While the coupled effects of heat and pressure, and welding speed, are broadly understood today as keys to a successful ERW seam, it is generally not understood that their coupled significance has been recognized and written about for almost 90 years. Their coupled effects are elaborated in the patents [4,5] that underlay the first successful production of an ERW-seamed product, which was short-length small diameter tubes. It is more clearly evident in the written record of the early patent infringement litigation, which occurred between then competing ERW tube-making processes, with those three factors becoming the benchmark for such decisions [6]. In this specific context, there appears to be little practical difference between the success factors for producing small diameter decorative tubes in the early 1920s versus those for large diameter pressure pipe production over the period of what is now almost a century.

Over the years, the various applicable API Specifications for line pipes[1] have termed long ‘seams joined electrically welded without the addition of extraneous metal’ – which by definition is an autogenous weld – as ‘electric weld’ pipe. The API usage of this ‘electric weld’ terminology since the 1940s is perhaps a consequence of a search then of the history of the ERW process, which then as now leads to a patent granted to Thomson in 1886 [8]. The apparatus and claims of that patent define a method involving an autogenous upset butt weld between the ends of two pieces of wire. The text of the introductory paragraphs to that 1886 patent describes the process that over time has become known as electric (resistance) welding, which Thomson initially coined ‘electric welding’.

Rotating Electrode (LF) ERW Processes

Subsequent patents adapted that early electric upset welding practice to close the long seam in tubes, and later in pipes – for which the term tube as used in the 1920s represented a small-diameter circular cross-section of the order of a few inches or so. The first of the patents to adapt the electric weld concept (and today’s ERW process) appears in 1889, which also was granted to Thomson for an electric lap weld [9]. Little of consequence in making tube and/or pipe emerged until 1898 [10], when what appears to be the first patent to close a butt-welded long-seam in a tube using electric welding was granted. Apparently impractical based on consideration of its method [e.g., see 11], that patent was followed just two years later by the patent referred to in the subsequent extensive infringement litigation as the Parpart Process. [12] As becomes evident current was initially delivered to the skelp through wheel type electrodes that were supplied by an alternating current (AC) source. Such seams have been termed low frequency (LF) electric resistance welds (LF ERW). Frequencies for such welds are often cited at up to 360 Hz, although values as high as 900 Hz also have been reported [6]. Because of the low frequency and the wheel contact scheme adopted, the resulting ERW seams had a comparatively wide heat-affected zone (HAZ).

A comparable patent of consequence to making tube and/or pipe was granted in Europe in 1912 [13], whose claims also reflect a method for closing of the longitudinal seam of tubes by electric resistance welding. The figures and loosely translated claims of that Austrian Patent, as amended in its filing in Germany, indicate that two rolling electrodes are coupled with two lateral ‘pressure’ rollers to close the seam along a previously formed cylindrical can. The images in the 1912 patent show the position and arrangement of the roller electrodes, and the shape and confinement provided by the lateral pressure rollers. All such aspects are comparable to the related aspects shown in the many figures in the two above-noted Johnston Patents [4,5], which were granted much later in the US (in 1921 and 1922, respectively). The similarity between these method patents is evident in Figure 1.

   

a) from the 1912 European rotating electrode patent [13]

b) concept for the 1922 Johnston patent [5]  (the image is from Reference 14)

Figure 1. Similarities in concept between early patents filed in Europe and the US

The view shown in Figure 1a is reproduced from the afore-noted 1912 German patent [13], while the view in Figure 1b shows the setup of the Johnston patents, which is reproduced from a 1934 Handbook [14] on tube production published by Steel and Tube, Inc., (S&T) to document their process and apparently also market their products. At the time that handbook was published, S&T was held by the Elyria Iron and Steel Company (EI&SC), which was assigned the rights to the second of the Johnston Patents [5]. In turn, EI&SC became part of the Republic Steel Corporation (RSC), as it was formed circa 1930. While the level of detail that exists between the renderings shown in Figure 1 differs slightly, this difference is consistent with trends in the patent literature over the decade that has passed between these images. While the components were not identified in the original patent drawing, the view in Figure 1a includes the centerlines of the pressure rollers, which are shown there only in part. It is apparent from Figure 1a that a pair of rotary electrodes lie above the pipe, which from their shape and position in this elevation also serve to confine the pipe’s shape. Comparison of Figure 1b with that in Figure 1a indicates that little difference exists in concept: the only essential difference lies in the addition of a lower support roll for the early 1920s concept.

Both Johnston patents [4.5] are broadly referred to in US infringement litigation associated with early ERW production. Litigation [6] that involved S&T as the infringed party, which makes clear that the Johnston patents underlie the first successful ERW long seam process in the US, begins in the mid-1920s and continues well into the 1930s [e.g.,6,11,15,16]. Such litigation is important in the present context because it establishes the timeline for the capability to make a functionally sound ERW seam using a rotating electrode process as being the early to mid-1920s. According to claims in Reference 14, at the time of its publication tube was being commercially produced in accordance with the Johnston patents in diameters up to 4 inches, and in wall thicknesses up to 0.238 inch. Tables in Reference 14 cite pressure capacities in excess of 4000 pounds, while other aspects note the use of flattening, cross-seam tensile testing, and pressure testing as part of the product’s functional qualification. While it is clear that the initial production involved smaller diameter thinner-walled products, it is equally clear that products made circa the early 1930s involved high pressure applications in diameters up to 4 inches. The image of the through-seam cross-section reproduced from Reference 14 here in Figure 2 is not inconsistent with such claims. As shown therein, the upset material is not removed in that image – although various approaches to process the ‘flash’ as it was called were offered. Finally, it is noted that “because scrap increases at a greater proportion in length” they offered the product in a standard length of 16 feet.

Figure 2. Cross-section montage through an ERW seam produced prior to 1934 [14]

Within a year of filing his initial (process) patent, Johnston applied for a patent on an apparatus to electrically butt weld tubing, which while applied for in 1920, was not granted until 1924 [17]. While that tube-machine patent was still pending, Johnston applied for the tubular product patent that was cited above as Reference 5. Shortly thereafter, Johnston also recognized the need for and had developed approaches to improve his seam-process patent, with the related application filed in 1923 [18]. In parallel, EI&SC (who held the rights to the Johnston patents for tubular products and the apparatus) filed a quite comprehensive process and product patent in Great Britain in 1924 [19]. Two years after the Johnston ‘apparatus’ patent application and even longer after his initial process application, Axel and Nels Johnson [20] filed a patent that claimed improvements to the Johnston process. While the 1922 application of Axel and Nels Johnson was granted in 1924, as was that for Johnston’s apparatus based on 1920 application, the earlier application controlled in the related infringement litigation. Even so, the similarity in names involved opens to the somewhat confusing the history that underlies the rotating electrode ERW long seam process.

In summary, the 1922 Johnston patent establishes the coupled relationship between current (heat), upset pressure, and welding speed, which served as the benchmark to establish those patents as the basis for the first successful ERW seam production. The Johnston patent two years later eventually paved the way to economic production of products with a heavier wall than the tubing of that time, and so opened the door to produce heavier wall ERW line pipe.

The Rotary Electrode Processes is Scaled-Up for Line Pipe Production

The first easily located process patent located that included the terms ‘electric welding’ and ‘pipe’ in its title was granted in 1934 to Blevins [21][2], and assigned to RSC. This patent made use of the rotating electrode concept, possibly because this was the basis for the Johnston tube process held by S&T (which then also was part of the RSC).

The Blevins patent is a key benchmark in the timeline for the rotating electrode process: it establishes that the maximum wall thickness heretofore produced commercially at ~0.125 inches (3.18mm) and a maximum diameter of ~four inches (102mm). That Blevins assigned his patent to the RSC suggests that he would be aware of the scale of the S&T production, such that it is not a surprise that the diameter he cites is consistent with the maximum size claimed by S&T. While this serves to benchmark the maximum diameter produced circa 1934, the maximum thickness cited in this patent is much less than that claimed by S&T in their 1934 handbook [14], which leaves the maximum thickness produced uncertain. However, this difference could also reflect a change due to recent S&T developments that had not yet been conveyed to Blevins.

Figure 3 illustrates the key features of the Blevins patent, with this view being cut from a much more complex view of the apparatus. Specifically, the view shown is cut from near the shaft centerline of the rotary electrodes down through near the shaft centerline of the lower support roller, which encompasses the essential features of the method.

Figure 3. Pressure rollers and rotating electrodes circa 1935 [21]

When compared to the image for the Johnston patent of about a decade prior, or the German patent more than two decades earlier, it is clear that the level of detail evident between these renderings differs greatly. This is not an indication of differences in process complexity, but simply rather reflects the trend to present more detail, which is apparent in the patent literature over that period.

Although a decade had passed, the position and arrangement of the pressure rollers and the rotary electrodes for Blevins’ patent are comparable to those shown in Figure 1b. Figure 3 indicates a pair of rotary electrodes, identified as item 18, set above the pipe which from their disposition in this elevation and the related text also serve to confine the pipe shape. They act together with a pair of lateral ‘pressure’ rollers, identified as item 105 and its unnoted partner, and a lower support roller, just as in Figure 1b. As such, the major aspects of the process are similar in concept over the decade that falls between these patents, with similar circumstances being the case in the vicinity of the welding station throughout LF ERW production using rotary electrodes.

Key aspects of this rotary electrode LF ERW seam process changed little since the concept was first practiced at a smaller scale to make tube since the mid-1920s, with the process as patented in Europe circa 1912 also being similar. Such practices were used into the late 1960s, but began to be abandoned as the more efficient HF ERW process became broadly commercial in the US into the early 1970s.

A co-pending application to that of Blevins (also assigned to RSC) was filed by Neckerman in 1931 that made reference to the ‘continuous production’ of electrically welded pipe of large diameters and heavy gauge material. That patent, granted in 1935(23), also is an important benchmark in time for the rotating electrode process, because it makes clear that the so-called ‘continuous production’ was still a can-by-can process. In contrast to the inference, based on the patent the reference to ‘continuous’ was specific to provision of a system that carried from inbound skelp to outbound pipe. Based on the patents, truly continuous production was not considered until into the 1950s, as becomes evident later.

While patents are seldom in hand prior to the first production, it is usual to apply for patent protection at a point in time that the process is reduced to practice and the product is in hand. On that basis, it is unlikely that larger diameter or heavier wall pipe was being produced on a commercial basis using a rotating electrode process before 1931. Finally, while no constraints on pipe size are indicated, reference is made to current demand(23), among other factors, that led to noting a thickness of 9/16 inch and a diameter of 16 inches. On that basis one could infer, at least over the near term circa 1931, that sizes larger than that were not commercially produced.

The Flash Weld Process

At the same time that the rotary electrode process was being developed to make tubing and pipe, AOS was contemplating a ‘flash’ weld type of ERW process to make tubing. In co-pending applications for process patents in 1928 [24,25], AOS set out to make tubing and pipe by making the long seam weld over the full length of a preformed can ‘substantially immediately’ in a simple and efficient apparatus. Direct current was supplied to the edges to be joined in that apparatus, with the patents granted in 1932 [24,25]. Figure 4 provides one view of the FW process as it was characterized in the early 1930s. This figure shows just the portion of the pipe-making machine that held the pre-formed can, and closed and welded the seam. In this view, items 58 and 41 are linear electrodes that ran the length of the pipe. The pipe is centered in this view, from which it is apparent that the pipe is supported on its exterior, but also supported on the interior local to the loading due to the linear electrodes.

Apparently also concerned for issues related to the control and other aspects of the rotary electrode process, AOS was granted a patent in 1930 for this apparatus, which included the claim that it was ‘simple and efficient’ to operate that would create the long seam simultaneously over its length based on a process patent applied for in 1930 [26].

Figure 4. Patent-based view of a cross-section through an AOS machine circa 1932 [25]

Consistent with the earlier assertion that patents are applied for as a process is reduced to practice, the AOS website [27] notes that in 1927 it perfected a method to economically form and weld and so mass produce large-diameter steel line pipe. While it cannot be confirmed from the patents or other literature, it appears that the processes referred to in reference to 1927 involved the 1928 patent applications. In spite of the initial patent applications being made circa 1928, a 1928 newspaper [28] notes that production of 400 miles (of unstated diameter pipe) was underway by AOS, and also indicated that they then could produce pipe in diameters up to 26 inches in segments 30 feet in length.

It follows in view of the above discussion that large diameter FW pipe was produced and likely went into service prior to pipe produced using the rotary electrode concept.[3]

Evolution of the Early ERW Processes – A DC-Based Rotating-Wheel Process

The FW process just discussed relied on a DC source, and created the seam ‘simultaneously’ over the length of the pipe, with a suite of patents developed to cover that process [24-26]. In contrast, the rotary electrode process adopted by the RSC closed the seam ‘progressively’ from one end of the pipe to the other using alternating current [e.g.,4,5,17,18,29]. Because each of these two much different schemes had established process patents, either alternative schemes had to evolve as a ‘design around’ the concepts and claims of the existing rotary electrode and simultaneous FW patents, or the user had to pay in some way to make use of these patented processes. Accordingly, it is not a surprise that the 1930s gave rise to alternative pipe making practices that either changed or broadened the utility of the existing schemes. The new schemes had to differ in some important way to avoid the infringement litigation, which had been initiated in the mid-1920s in regard to the Johnston patents and remained active for over a decade. Key among those early developments included the DC welding scheme developed by the Direct Current Welding Company (DCWC), and by the Youngstown Sheet and Tube Company (YS&T) [30,31], which also closed the seam ‘progressively’ from one end of the pipe but used direct current.

Key considerations for all such patents included the power level required, the efficiency, the type and disposition of the transformer, and the approach used to supply the current to limit the losses.

The above-cited documentation establishes the RSC as the first successful rotary electrode pipe maker, circa the early 1930s [21], and establishes AOS as the first large-diameter pipe maker, circa 1928 [22], with Reference 27 inferring the possibility it was made as early as 1927. The RSC practice is termed today LF ERW, whereas the AOS practice is today termed flash welding, as it was in the 1920s. While the process names differ, both create the long seam using an ERW process. Without reference to such documented history, the National Institute of Standards and Technology (NIST) states in their 1989 report on ERW seam failures [32] that ERW “pipe” was first produced 1929. Kiefner and Clark [33] also note a timeline for the first ERW production with reference to a 1957 paper [34] that considered this history. Reference 34 states that in “1933 production of ERW pipe was begun …..” – which is close to the date of the Republic patent for large-diameter pipe production(22), but well after 1929. Kiefner and Kolovitch [3] (without citation) subsequently noted the first production occurred in 1929, and indicate (without citation) that the RSC was the first to produce pipe with an ERW long seam.

Pipe Making Began Joint-by-Joint, and Continued So Apparently for Decades

Regardless of the specific date or the first producer, suffice it here to note that ERW pipe was commercially available by about 1930. Review of the early tube-scale production patents indicates that a major change occurred in the process sequence about the late 1920s. Up until then, tube was made from precut lengths of skelp that were preformed into a tube shape and then welded. Starting in the late 1920s, production on a tube-by-tube basis shifted to a truly continuous process, with the tubes cut to length after welding. [35,36]. Study of the patents for early pipe production [21,37,38] makes clear that this production process closed the long seam in preformed cylinders, termed a ‘blank’ or a ‘shell’, as had been the case early on for tubes. Figure 5 reproduces a view of one of the early joint-by-joint production sequences, circa the mid-1930s.

In some cases, the patents for the early processes also indicate that preheating was used apparently to assist in the forming process, as well as to facilitate the welding process – even though this would form an oxide skin that in extreme cases might compromise the weld quality.

Figure 5. View of one early (circa 1934) joint-by-joint production sequence(23)

Joint-by-joint production opens to setup and process variations, in contrast to what might be achieved using a steady-state process that made pipe from a continuous supply of skelp, which would separate the completed pipe from the continuous process after the long seam was closed. In spite of the benefits that could accrue to a continuous process, apparently due to complications in that approach and other factors, patent applications for joint-by-joint pipe production continued in the 1940s, and on into the 1950s[4]. Patents for such processes were filed in that timeframe [e.g.,39,40], with some as late as 1956 [41]. That said, patents for truly continuous ‘progressive’ electric resistance seam welding processes [42] also were being filed, such that by the 1950s pipe was being made with ERW seams joint-by-joint, as well as by continuous processes. Of significance in regard to the continuous processes is that some patents [e.g.,43] sought to limit the temperature of the preheat step. While advantageous as a cost saving and to minimize bondline oxides, eliminating the preheat also opened to the possibility of more rapid cooling of the seam, and the potential for untempered martensite in some cases. Such a transition in process could explain in part the relatively high incidence of failures in LF ERW pipe produced in the 1950s timeframe[5].

Summary for the LF Processes

It follows from the above discussion that the LF ERW processes have changed little in concept since the late 1920s, with the essential differences over time being driven by the need to scale up in size and/or improve efficiency. It also is clear from the discussion in the 1934 S&T Handbook [14] that the process as it was practiced in that era could produce a ‘quality’ seam. That Handbook makes clear that seams of that era survived a range of testing, including flattening, cross-weld tensile tests, and burst tests, all of which showed failure remote to the seam. While the steels of the 1930s era have since been much improved in terms of cleanliness/chemistry and processing, some grades were listed in the Handbook with values of the ultimate tensile stress (UTS) up to 80 ksi. Given that those steels failed in burst testing remote to the seam, the LF process of the early 1930s could produce a very strong viable seam that was fit-for-purpose, which remained the case up through the transition to the HF processes.

It also follows that the issues historically evident for the LF processes do not imply that those processes were inherently problematic, but rather suggest that upsets in those processes underlie the in-service and hydrotest failures. Finally, it follows that until upsets in such processes are managed (i.e., avoided or detected online), such in-service failures can be anticipated to persist.

Practical HF ERW and HFI Processes Emerge Circa the Late 1950s

As the processes collectively termed LF ERW evolved, it became evident that much more efficient production could be achieved via high-frequency (HF) processes. This was because the HF processes could focus the energy to create the bondline between and local to the interfaces to be joined. This process efficiency and other potential benefits due to the use of a HF in lieu of a LF ERW process were recognized quite early. For example, a 1931 patent was filed that while granted much later in 1937 [44] targeted ‘improvements in the electrical methods of heating plates or tubes’. The introduction to that patent states “my invention is a method of confining or concentrating the heating current to the portions in which the heating is desired.” It continues noting that the benefit “is to lessen the cost and improve the quality and uniformity of the welding by reason of superior control.” While the introduction to that patent only vaguely notes “using an alternating current of the necessary frequency,” later in that patent reference is made to a frequency the order of 30 KHz. It was, however, until the 1950s that the vision of a contact-based HF ERW concept from the 1930s found practical utility through a series of patents due to Rudd, and coworkers [e.g., 45-47]. Likewise, it was about the 1950s that non-contact HF ERW – so called HFI – began the transition from concept into practice [e.g., 48]. Rudd and his colleagues were likewise engaged in this aspect [e.g., 49], along with others early during its development [e.g.,50]. These early HF patents laid the foundation for the developments that continue to evolve today, with current production based on this concept using what could be viewed as third-generation technology relative to that at the close of the 1960s.

As for the LF processes, little has changed conceptually following the early developments that ran into the later 1970s, with the essential differences over time being driven by the desire for better control and improved efficiency. Research has targeted theoretical modeling and understanding of the physics [e.g.,51], and the capability to ‘tune’ both the process and mechanical setup to produce a given thickness of skelp [e.g.,52,53]. In addition, technology has become increasingly available as the years have passed, which has been adapted to sense temperature, and other process variables. References 52 and 53 are good examples of the adaptation of technology to understand the effects of heating, while such adaptation in the late 1970s capitalized on advanced image processing and related aspects.

Summary

Both contact and noncontact HF processes have benefited from technology development or adaptation, in regard to process control [e.g.,54,55], and online inspection [e.g.,54-56]. Each has also benefited from the availability of skelp that has been designed specifically for such applications [e.g.,57], although this and the other benefits begin to become evident primarily for the post-1980s production. The related literature indicates that aspects of the technologies involved to date capitalize on adaptation from parallel fields, as well as make use of developments specific to ERW production.

As for the LF processes, work done to evaluate the strength and other aspects of the quality of HF seams indicates that when the seam is produced under control it has properties that are comparable to the pipe body, and otherwise is free of integrity issues [58-60]. Thus, failures for pipes with a HF seam do not indicate that the HF processes are problematic. Rather, as noted for the LF processes, such failures reflect upsets in the process setup and/or production of the seam, and in the supporting quality controls and quality assurance practices. And as for the LF processes, such failures also can be due to issues with skelp quality – whether the seam is made in the early 1960s or post 2000.

It follows that until upsets in the HF processes that have the potential to fail in-service are uniformly managed (i.e., avoided or detected online), failures can be anticipated to occur. Because both the contact and noncontact HF processes have benefited from technology development or adaptation in regard to process control [e.g.,54,55] and on-line inspection [e.g.,54-56], it is reasonable to anticipate that the HF processes will be inherently less prone to failure than the LF processes. The trends evident in Figure 1 bear this out. Such controls must ensure adequate heat supply and cross-seam upset to create the weld, with appropriate inspection available to detect the consequences of upsets when the controls fail across the pipe for all producers sourced by US operators. The implications of these topics are discussed further in the QR Segment Compare-Contrast LF and HF Processes.

References

  1. Messler, R. W., Principles of Welding: Processes, Physics, Chemistry, and Metallurgy, Wiley-VCH, 1999.
  2. Leis, B. N. and Nestleroth, J. B., “Battelle’s Experience with ERW and Flash Weld Seam Failures: Causes and Implications,” Battelle Interim Report, Subtask 1.4, US DoT Contract No. DTPH56-11-T-000003, September, 2013
  3. Kiefner, J. F. and Kolovitch, K. M., “ERW and Flash Weld Seam Failures,” KAI Interim Report on Task 1.4, , US DoT Contract No. DTPH56-11-T-000003, September, 2012
  4. Johnston, G. V., “Method and Apparatus For Butt Welding Thin Gage Tubing,” US Patent 1,388,434, 23 August, 1921.
  5. Johnston, G. V., (assigned to Elyria Iron & Steel Co) “Butt Welded Thin Walled Tubing,” US Patent 1,435,306, 14 November, 1922.
  6. Steel and Tubes, Inc. V. Greenpoint Metallic Bed Co, before District Judge Galston, 37 F.2d 172, 31 December, 1929
  7. Anon., Specification for Line Pipe: API Specification 5L, American Petroleum Institute, Dallas, 2010.
  8. Thomson, E., “Apparatus for Electric Welding,” US Patent 347,140, 10 August, 1886.
  9. Thomson, E., “Method of Electric Welding,” US Patent 444,928, Jan. 20, 1891.
  10. Ries, E. E., “Method of and Means for Manufacturing Metal Tubing,” US Patent 611,222, 20 September, 1898
  11. Steel & Tubes, Inc. V. General Tube Co., Circuit Court of Appeals, Third Circuit, Nos. 4750 and 4844, before Circuit Judge Davis, 3 October, 1932.
  12. Parpart, O., “Electric Welding Machine,” US Patent 658,741, 25 September, 1900.
  13. Gesellschaft fuer Elektrotechnische Industrie “A Method for Closing the Longitudinal Seam of Tubes using Electrical Resistance Welding,” German Patent AT 58,866D, 15 November 1912.
  14. Anon., Handbook of Electric-Weld Tubing, Steel and Tubes Inc, 1934
  15. Elyria Iron & Steel Co. V. Mohegan Tube Co., before Circuit Judges Hough, Manton, and Hand, 7 F.2d 827 and 7 F.2d 829, 21 May, 1925
  16. National Tube Co. V. Steel & Tubes, Circuit Court of Appeals, Third Circuit, before District Judges Buffington, Davis, and Dickinson, 6 May, 1937
  17. Johnston, G. V., (assigned to Elyria Iron & Steel Co) “Apparatus for Electrically Butt-Welding Tubing,” US Patent 1.501,782, 15 July, 1924.
  18. Johnston, G. V., “Electric Tube Welding Machine,” US Patent 1,652,481, 13 December, 1927.
  19. Elyria Iron and Steel Company, “Improvements In and Relating to Electric Welding,” British Patent No. GB 216,203A, 21 May 1924.
  20. Johnson, A. Z. and Johnson, N. A., “Apparatus for Butt-Welding Tubing,” US Patent 1,492,173, 29 April, 1924.
  21. Blevins, H. G. (assigned to Republic Steel), “Method and Apparatus for Electrically Welding Pipe,” US Patent 1,982,415, 27 November, 1934.
  22. Scheil, M. A., Fratcher, C. E., Henry, S. L., Uecker, E. H., “Manufacture and Metallurgy of Flash-Welded Line Pipe,” The Engineering Journal, February, 1957, pp. 60-71.
  23. Neckerman, W. M. (assigned to Republic Steel), “Apparatus for Manufacturing Electrically Welded Tubing,” US Patent 2,023,897, 10 December, 1935.
  24. Crawford, W. E., (A O Smith) “Electrical Welding Apparatus,” US Patent 1,857,115, 3 May, 1932.
  25. Andren, B. T., Hineman W. F., and Crawford, W. E., (A O Smith) “Welding Tubular Articles,” US Patent 1,872,055, 16 August, 1932.
  26. Crawford, W. E., (A O Smith) “Method of Electrically Welding Tubular Articles,” US Patent 1,783,411, 2 December, 1930.
  27. Anon., A O Smith Corporate Website, www.aosmith.com /About/Detail.aspx?id=130 accessed 7/4/2012.
  28. Anon., Logansport Pharos Tribune, May 18, 1928, Page 4 http://newspaperarchive.com/logansport-pharos-tribune/1928-05-18/page-4, accessed 7/4/2012.
  29. Kurtze, R. C. F., (assigned to Republic Steel) “Apparatus for Electrically Welding Tubing,” US Patent l,998,l39, 16 April, 1935.
  30. Youngstown Sheet and Tube, “Improvements In or Relating to Welding Methods and Apparatus,” United Kingdom Patent GB369,431, 24 March, 1932
  31. Mirfield, G. E., Lewis, R. E, Firth, C., and Neale, J. E., (assigned to Youngstown Sheet and Tube) “Electrical Welding Apparatus,” US Patent 2,003,118, 28 May, 1935.
  32. Fields, R. J., Pugh, E. N., Read, D. T., and Smith J. H., “An Assessment of the Performance and Reliability of Old ERW Pipelines,” NISTIR 89-4136, National Institute of Standards and Technology, 1989
  33. Kiefner, J. F. and Clark, E. B., History of Line Pipe Manufacturing in North America, ASME CRTD-Vol 43, 1996.
  34. Thomas, P. D., “Pipe – Manufacture and Use,” Iron and Steel Engineer, July, 1957, pp. 84-87.
  35. Sessions, F. L., (assigned to Elyria Iron & Steel Co), “Flying Shears,” US Patent 1,648,829, 8 November, 1927.
  36. Johnsen, B., (assigned to Steel and Tubes, Inc), “Cut-Off and Delivery Mechanism,” US Patent 1,830,160, 10 November, 1931.
  37. Sessions, F. L., “Welding Machine,” US Patent 1,808,261, 2 June, 1931.
  38. Blevins, H. G., (assigned to Spang Chalfont) “Apparatus for Welding,” US Patent 2,084,889, 22 June, 1937.
  39. Fleche, J. R., (assigned to Republic Steel) “Method of Progressively Electric Resistance Butt Welding Longitudinal Seam Edges of a Tubular Blank and Apparatus There For,” US Patent 2,417,594, 18 March, 1944.
  40. Babbit, C. A., (assigned to Cal-Metal Corporation) “Pipe Welding Machine,” US Patent 2,613,303, 7 October, 1952.
  41. McDowall, W. S., (assigned to Republic Steel) “Electric Weld Tube Mills Inside Spreading Rolls at Electrode,” US Patent 2,816,208, 10 December, 1957.
  42. Colby, I. A., (assigned to Youngstown Sheet and Tube) “Method of and Apparatus for Continuous Welding of Pipes and Tubes,” US Patent 2,575,381, 20 November, 1951.
  43. Clark, E. B., Leis, B. N., and Eiber, R. J., “Annex E in Integrity of Vintage Pipelines,” Battelle Final Report to INGAA Foundation, October 2004.
  44. Bennett, E., (assigned to Wisconsin Alumni Research Foundation), “Method of Heating Electrically Conducting Bodies for Industrial Operations,” US Patent 2,066,668, 5 January, 1937.
  45. Rudd, W. C. and Stanton, R. J., (assigned to The Magnetic Heating Corporation), “Seam Welding by High Frequency Resistance Heating,” US Patent 2,818,488, 31 December, 1957.
  46. Rudd, W. C. and Stanton, R. J., (assigned to The Magnetic Heating Corporation), “Tube Welding,” US Patent 2,857,503, 21 October, 1958.
  47. Rudd, W. C. and Stanton, R. J., (assigned to The Magnetic Heating Corporation), “Welding by High Frequency Resistance Heating,” US Patent 2,876,323, 3 March, 1959
  48. Body, A. C., (assigned to The Ohio Crankshaft Company), “High Frequency Induction-Heating Apparatus,” US Patent 2,654,019, 29 September, 1953.
  49. Rudd, W. C. and Stanton, R. J., (assigned to The Magnetic Heating Corporation), “Induction Welding,” US Patent 2,763,756, 18 September, 1956
  50. Kohler, F., (assigned to The American Machine and Foundry Company), “Method and Apparatus for the Induction Welding of Tubing,” US Patent 3,037,105, 29 May, 1962
  51. Scott, P., “High Frequency Welding of Low Carbon Steel Tube,” Thermatool Corp Paper, undated, www.thermatool.com/…/welding/ High Frequency Welding o….
  52. Haga, H., Aoki, K., and Sato, T., “Welding Phenomena and Welding Mechanisms in High Frequency Electric Resistance Welding – 1st Report,” presented at the AWS 60th Annual Meeting Detroit, 2-6 April 1979, also in Welding Research Supplement, July 1979, pp 208s – 212s.
  53. Kim, D., Kim, T., Park, Y. W., Sung, K., Kang, M., Kim , C., Lee, C. and Rhee, S., “Estimation of Weld Quality in High-Frequency Electric Resistance Welding, with Image Processing,” Welding Journal, Research Section, March 2007, pp 71-S to 79-S.
  54. Koch, F. O. and Peters, P. A., “Distinguishing Characteristics of High-Frequency Induction-Welded Pipe,” presented at API Standardization Conference, Pipe Symposium, New Orleans, 23 June 1986.
  55. Baralla, E. and Tommasi, C., “Integrated System for Process Control of High Frequency Electric Resistance Welded Steel Pipe,” Proceedings PAN NDT Meeting, Rio de Janeiro, 2003
  56. Anon., “Complete Inspection Solutions for ERW Tubes (including Large Diameter Pipe),” Olympus Document, see http://www.olympus-ims.com/en/erw/, accessed July 2011.
  1. * This discussion borrows heavily from Leis, B. N., “Time-Trending and Like-Similar Analysis for ERW-Seam Failures,” Task 4.2 Contract DTPH56-11-T-000003, Pipeline and Hazardous Materials Safety Administration, July 2013.

    1 The API 5L Specification [7], which has existed since 1928, makes reference to ‘electric welding’ as it was called then first in the 5th edition, with that process defined later in the 8th edition (by supplement in late 1942). API 5LX Specification, which came into use in 1948 and remained in use through 1982, also addressed ‘electric welded pipe’ and ‘flash-welded’ pipe beginning in 1948.

  2. Given that AOS had produced large diameter ERW pipe in 1928 via the FW process [22], whether for commercial or other reasons apparently Blevins and the RSC did not consider the FW process to be an ERW process.

  3. According to the AOS authors that wrote Reference 22, the 1928 FW mill was producing pipe that was 40-foot in length, while the 1927 SMAW mill was not converted to produce FW pipe in 30-foot lengths until 1930. As such it appears that the pipe referred to in Reference 28 was not ERW pipe. Nevertheless, it is clear from Reference 22 that AOS was making large diameter FW pipe in 1928, such that this conclusion remains valid.

  4. Given 1) the capital sunk in equipment to produce joint by joint, and 2) the benefits by the transition to use of the emerging HF technology, it is reasonable to assume that companies that had a major investment in joint by joint production would persist in its use until HF processes were being perfected in the 1950s.

  5. Reference 43 summarizes ERW seam failures tabulated from various sources including Reference 33, with the data therein and the related analysis indicating the clustering of incidents in the 1950s, which could trace in part to the reduced use of preheating and/or a reduced preheat temperature.