An Assessment of Common Core Assembly Methods for Thin Gauge Non-Oriented

Electrical Steels

Steve Sprague Proto Laminations, Inc. 13666 East Bora Drive

Santa Fe Springs, California 90670 USA

This paper presents an overview of three core assembly technologies commonly employed in the manufacture of electrical machines made with thin-gauge non-oriented electrical steels: welding, adhesive bonding and in-die interlocking. This author presented a limited review of stacking technologies in a presentation given at WMM ’14. [1]; the following discussion will be extended here in greater detail and also with a review of technical and market developments. Introduction Whether manufactured by stamping, laser cutting, wire EDM machining or by other methods, electrical laminations must be assembled, stacked, into finished cores ready for winding and other subsequent manufacturing operations. Though these assembly processes may be used to produce stacks made from thin grain-oriented steels, the topics covered in this paper will pertain directly to non-oriented steels; the reader is cautioned to confirm their applicability to the use of grain-oriented lamination materials. Contemporary electrical machine design often calls for motors or generators to operate in regimens that increase the core losses of the steel used in the laminations, whether from higher rotational speeds, elevated excitation frequencies, or high pole switching frequencies, or many times from a combination of these factors. An important design strategy for improving the loss profile of such machines is to incorporate thinner laminations, often in the range of thicknesses less than 0.36 mm (0.014 inch), the class of lamination materials that have come to be called thin-gauge electrical steels. Others have noted that as lamination thickness is reduced, the complexity of lamination manufacture and assembly becomes ever more challenging. [2] This sentiment holds true not only for the manufacture of the individual laminations but also for the cores assembled from them. Thin-Gauge Electrical Steels Several types of electrical lamination materials are produced in thicknesses below 0.36 mm (0.014 inch), notably selected grades of grain-oriented silicon steels, nickel-iron and cobalt-iron alloys as well as the special class of 6½% silicon steels, the term Thin-Gauge Electrical Steel has developed over the past several years to describe non-

oriented silicon steels with thicknesses below 0.36 mm (0.014 inch) and with a silicon content in the 2.5% to 3.0% range. While several papers studying these electrical steels have been presented at previous meetings of the International Conference on Magnetism and Metallurgy [1][3][4][5], a brief reintroduction to their range of thicknesses, grades and specifications will help the reader organize these steels in the context of all non-oriented silicon steels and also in regard to the discussion of assembly methods that follows. Thin-gauge silicon steels are governed by several international standards, notably IEC 60404-8-8 [6] and ASTM A1086-13 [7]. These standards have established grades sorted by maximum core loss values. It should be noted that IEC 60404-8-8 is currently under review with a thorough revision, incorporating additional thicknesses and grades, expected to be completed by the end of 2016; additionally, a review of ASTM A1086 will begin in 2018. IEC 60404-8-8 edition 1 [6] — Maximum specific total loss tested at 1.0 T and noted frequency

medium to high frequency Epstein frame test Thickness Grade 400Hz 1000Hz mm (in) W/kg 0.05 (0.002) NO-5 45 0.10 (0.004) NO-10 13 0.15 (0.006) NO-15 14 0.20 (0.008) NO-20 15

ASTM A1086-13 [7] — Maximum specific core loss tested at 1.0 T and both 400 Hz and 1000 Hz,

medium to high frequency Epstein frame test Thickness Core Loss Type 400Hz 1000Hz in (mm) W/lb (W/kg) 0.004 (0.10) 10T590 5.90 (13.0) 12.0 (26.4) 0.005 (0.12) 12T610 6.10 (13.5) 15.3 (33.7) 0.006 (0.15) 15T640 6.40 (14.0) 17.4 (38.3) 0.007 (0.18) 18T650 6.50 (14.4) 19.5 (43.0) 0.008 (0.20) 20T680 6.80 (15.0) 22.4 (49.4) 0.009 (0.22) 22T700 7.00 (15.4) 25.0 (55.1) 0.010 (0.25) 25T730 7.30 (16.0) 28.0 (61.7) 0.011 (0.27) 27T770 7.70 (17.0) 30.2 (66.6) 0.012 (0.30) 30T820 8.20 (18.0) 32.8 (72.3)

All electrical steels less than 0.36 mm (0.014 inch) in thickness are termed Thin-gauge electrical steels, yet there are within this collection a comparatively wide range of thicknesses that fall generally into three categories: those from 0.30 to 0.25 mm (0.012 to 0.010 inch), from 0.23 to 0.12 mm (0.009 to 0.005 inch) and 0.10 to 0.05 mm (0.004 to 0.002 inch). These informal groupings represent the typical end-use demands they are called on to fulfill as well as the technical challenges to be overcome in their production and in the manufacture of the laminations and assembled cores made from them.

Core Assembly Techniques for Thin-Gauge Electrical Steels There are many lamination stacking techniques that find some measure of use when assembling laminations of any thickness, and these, in the main, are also used in the assembly of cores made from thin-gauge electrical steels. The choice of core assembly technique will depend on a number of factors. Most critical in this evaluation are the anticipated level of production (ranging from initial research investigations through prototype development to modest or high quantity manufacture) and also the technical suitability of a particular assembly technique given the thickness of the laminations to be stacked. As noted above, there is a wide range of thicknesses in this class of lamination materials and some manufacturing techniques may not be appropriate for certain thicknesses. In addition, some stacking techniques may prove to be limited in their ability to provide stacks that will perform properly in certain applications. The three assembly methods to be discussed in this paper have each proven to be successful methods for assembling thin-gauge electrical steels when properly applied.

Welding The welding of assembled lamination stacks has been in common practice for decades. Whether using traditional TIG or MIG processes, or by laser or electron beam methods, welding can be a cost-effective method for producing robust lamination stacks. In relation to thin-gauge electrical steels, welding is often used because it can provide an immediately finished core and is suitable for use with most grades and thicknesses of thin electrical steels. While requiring the cutting of individual laminations by stamping, laser cutting or other means, as well as the construction of an accurate assembly fixture, laser cutting requires neither the curing time of adhesive bonding nor the chemical storage and handling concerns native to bonded core production. Issues Related to Machine Design and Magnetic Performance Unlike adhesive bonding, discussed below, that until recently has been considered to have negligible effects on the magnetic properties of a lamination stack, lamination welding has long been known to present a number of concerns related to the degradation of magnetic properties due to the high temperatures applied to the local area of the weldments. In addition, the continuous weld bead drawn down the edge of the lamination core can develop a path for electrical shorting of the stack. Numerous studies have examined and delineated these effects. Schoppa, Schneider and Wuppermann present findings that welded stacks will have increased core losses of 10 to 20% depending on polarization level for a test core with four weldments. [8] Arshad et al arrive at a similar result, expressed somewhat differently, with losses increasing from 0.5 – 1.0 % per weldment [9]; their core had 12 weld beads and thus a total loss penalty of about 12%. These studies examined cores built from lamination steel with a thickness of in .050 mm (.019 inch) and tested at 50 hertz; extrapolation of these findings for assessment of thin-gauge materials at higher frequencies may be problematic. Lamprecth, Hömme and Albrecht investigated the eddy current loss mechanism for welded specimens made from 0.36 mm (0.014 inch) electrical steel at 400 Hz. [10] While they did not develop their findings into a treatment of total core loss, their findings also indicate an increasing loss profile with additional weld beads. Krings et al have studied welded cores made from thin-gauge silicon steel, in this case 0.20 mm (0.008 inch) non-oriented electrical steel. [11] They studied cores with 4 weld beads at frequencies up to 200 hertz. Their presented results at 1.0 tesla show a loss increase at 50 hertz of nearly 35% and at 100 hertz of just less than 30% when compared to a non-welded reference sample. Both the Arshad and the Krings studies make reference to the width of the lamination backiron and to the overall size of the test specimen as having played an important though uncharacterized part in magnetic property degradation. Krings suggests that some of the increased loss may be attributed to the narrow width of their welded test specimens allowing edge-cutting effects to produce some of the overall loss penalty. None of these papers provided a description of the type of welding process employed nor an assessment of the quality of the weldments. Lamprecht et al do note, though, that they expect “A slightly bigger welding line will lead to a lower connector resistor and will therefore generate significant higher losses.” [12]

Clearly, the combination of the direct thermal effects of the welding process, the number of weldments and the physical dimensions of the laminations will all assume an important role in the design of a machine intended to be assembled by welding. Thermal effects of welding Welding introduces very high, if localized, temperatures into the electrical laminations in the area of the weld beads. Not only will this process affect the magnetic properties of the core, as described above, but can also impart thermally-induced deformations such as warping and other mechanical aberrations in the core. Schade et al developed a wide-ranging investigation of laser welding practices for lamination cores, discussing the effects of applied temperature not only on the weld beads and material shape, but also on thermal degradation of applied insulation coatings in the immediate area of the weld. [13] Efforts to mitigate these effects typically include the control of the weld head feed rate, as well as the choice of the welding process. Investigations into strategies to reduce the heat generated during the welding process, such as the pulsed laser method demonstrated by Vegli [14], show promise for thicker laminations but appear to offer no benefit as lamination thickness approaches 0.36 mm (0.014 inch) or thinner. One item that has become a common understanding when welding laminated cores of conventional lamination thickness is the importance given to the selection of insulation coating. The choice of suitable coating is equally of importance when welding thin-gauge laminations and the same concerns apply: laminations with an organic, or partially-organic, coreplate may experience cracks, bubbling or other imperfections arising from localized outgassing of the organic coating material through the weld bead. Beckley [15] states that occasional “blowholes” of this sort will likely not cause problems in the strength of the stack, but it has been the experience of this author that such outgassing effects can lead to cracks and other weld failures contributing to a weakening of the stack if the welds are systemically poor. Coreplate types without volatile components will generally not exhibit these imperfections, but may provide a lower level of electrical insulation and will generally be more abrasive on tooling components, thus shortening tool life in high production scenarios. [16] This final comment brings to light the importance of the skill and artfulness of the technician and system designer when developing, processing and monitoring the welding of laminated stacks. Of all the common assembly methods, welding places the greatest premium on the knowledge and experience of those directly involved. Limits to automation As a process, welding is inherently slow and is resistant to inclusion in the high speed stamping production of motor cores where a set of thin-gauge laminations for one laminated stack may be stamped and bundled for assembly in 3 minutes or less. The types of in-die welding processes one might see in industrial or automotive mechanical componentry has not been incorporated in lamination production. Laminations can be assembled into appropriate bundles within the tooling but must then be removed and queued for secondary welding operations. Once removed from the stamping tool, laminations can then be automatically welded using robotically controlled fixturing and welding. Automated welding systems can provide for greater economy of production since once welded they are immediately ready for subsequent operations: secondary machining, coating, winding or other manufacturing operations.

Adhesive Bonding Assembly of laminated cores using adhesive bonding techniques is a common practice, providing for stacks of sufficient strength to allow subsequent grinding or honing of the stacks to finished sizes, and to then proceed to coil winding and motor assembly operations. The general practice is to produce the laminations individually, regardless of cutting method, and then to assemble the individual laminations and bond them together as a subsequent operation. More recent advancements in automated adhesive application and curing provide the opportunity to investigate economical high production scenarios. Generally, any bonding process will include the following procedures: — the construction of a fixture that will hold the laminations in place during the bonding operation. — application of the adhesive. Depending on the adhesive system to be used, this can take place prior to the assembly of the stack or after the initially assembly where the laminations are loosely held and adhesive is applied and allowed to “wick” in to the core between the laminations. — careful assembly of the laminations onto the fixture. This will likely include the application of a “mold release” agent that will prevent the cured core stack from sticking to the fixture. — curing of the adhesive with controlled thermal processing. — careful removal of the bonded stack from the fixture. The bonded stack will then be ready for subsequent cleaning and other secondary machining and motor assembly operations. Stacking factor and other issues related to machine design and performance Adhesive bonding has often been considered to be a magnetically benign process, allowing laminations to be assembled without altering their magnetization and loss properties. Properly done, there is no contact between laminations when they are glued together; consequently, assembled cores do not experience the potential electrical shorting exhibited in welding, interlocking or other mechanical fastening schemes. However, the compression brought to bear on the lamination stack during fixturing and bonding may have an impact on the magnetic properties of the assembled laminations. Yamamoto and Yanase [17] and Myagi et al [18] describe the potential loss benefits and degradations of such applied compressive stress. It should be noted that these factors will likely be present in all core assembly methods where a compressive force is applied to assemble the laminations. Of potentially greater concern when working with bonded lamination stacks is the bondline thickness of the adhesive, typically in the range of 0.013 mm (.0005 inch) [19], and its contribution to lower iron density in the finished core. The ratio of the volume of electromagnetic material to overall core volume has come to be known as “stacking factor” among machine designers and machine manufacturers, yet this term yields some confusion when analyzing potential machine performance since the same term is used in a very specific way in the steel making and standards communities to assess

the quality of rolled steel. Some of this confusion is due to conflicting statements within the various standards themselves. — Stacking factor, more properly termed Lamination Factor in terminology libraries:

Standard ASTM A340 [20], provides a glossary of definition for magnetic properties and offers this definition of the term: Lamination Factor, (space factor, stacking factor), S—a numeric, less than unity and usually expressed as a percentage, which is defined as the ratio of the uniform solid height h of the magnetic material in a laminated core to the actual height h' (core buildup) when measured under a specified pressure. S is thus equal to the ratio of the volume of magnetic material in a uniform laminated core to the overall geometric volume of the core. This definition does not include the adhesive thickness or imperfections arising from lamination cutting and core assembly.

— Stacking factor, more properly termed Lamination Factor, in test specifications:

Standard ASTM A719/A719M [21], the test method for measuring Lamination Factor, provides a very specific definition of this term, differing from the one in A340: Lamination factor, S, indicates the deficiency of effective steel volume which is due to the presence of oxides, roughness, insulating coatings, and other conditions affecting the steel surface. One can see that with this second definition, the term directly refers to the quality of the as-rolled and as-supplied steel.

Since Lamination Factor or Stacking Factor are in common use with two potentially different meanings, one must be careful to ascertain the context in which these terms are used. When providing “stacking factor” values in product literature and data sheets, steel manufacturers rely on the type of meaning presented in A719/A719M and similar standards with stacking factor values often in the range of 95% and higher [22]. Should machine designers use the values presented in commercial data bulletins, while thinking that such values actually represent the “iron density” of the completed motor core, they will likely find that the percentage of magnetic lamination material in their core will be significantly lower than expected. This situation will be exacerbated when assembling cores of thin-gauge electrical steels. For a given adhesive system, the bondline, or thickness of the adhesive layer between laminations, will remain the same regardless of the thickness of the material. Insulative coatings applied at the producing mill, such as C-5 type coatings, will also have the same depth regardless of steel thickness. The diagrams below show the approximate variation in Stacking Factor as defined by ASTM A719/A719M and also as more loosely defined in ASTM A340 in an adhesive bonded stack.

Clearly, adhesive bonding affects machine performance in ways that must be understood at the design stage. Bondline thickness can vary to as much as 0.05 mm (0.002 inch), much greater than in the examples shown in these figures, thus consultation with the bonding technician in the early stages of design can help alleviate unexpected core density issues later on. Typically, thin-gauge non-oriented electrical steels are delivered with an insulative coating on both surfaces. However, this material may be supplied with coating on just one surface, or with no coating on either surface. Since the coating thickness makes a significant contribution to stacking factor, eliminating the coating from or both surfaces can provide one strategy to improve the “iron density” in the core, though this practice places extra pressure on the adhesive system to provide not only the bonded lamination structure but the entire interlaminar resistance as well. Thermal Considerations The adhesive systems used in the bonding of electrical laminations are cured at temperatures in the range of 177 °C (350 °F) [23] with curing cycles of up to 60 minutes depending on the exact requirements of the adhesive system. Once cured, though, these adhesives generally have a typical operating temperature of approximately 150 °C (300 °F) with peak ratings of about 205 °C (400 °F). The highest NEMA insulation rating commonly in use for electric motors, H, limits maximum temperature to 180 °C (356 °F). [24] Thin-gauge electrical steels are often used in machines where elevated

temperatures may be encountered, so the designer must carefully assess the potential thermal properties of the adhesive to be used. If the machine will have a likelihood of operating in the upper ranges of the insulation class, other lamination stacking methods that do not present such thermal limitations such as welded or interlocked stacks may be necessary. Advances in automation Lamination bonding is an inherently labor intensive process, with extensive handling and cleaning of laminations, as well as the construction of stacking fixtures, all contributing to a high labor content in the cost of the assembled core. Electrical motors and generators made from thin-gauge electrical steels have been traditionally thought of as a specialized class of machine, one that can absorb the additional labor cost of assembly without concern. However, as these materials have begun to find their way into higher production scenarios, manufacturers have sought ways to automate the assembly of their laminated cores. In those cases where interlocking is not possible due to the thin nature of the lamination steel, or where designers do not want to accommodate the potential performance degradation that may arise from welding, interlocking, or other schemes, two production methods may eliminate some or all of the assembly labor and cost. One is the use of laminations that are pre-coated with adhesive instead of, or in addition to, conventional insulation coatings. This adhesive-coated steel is designed to withstand stamping processes and can produce punched laminations that are ready to assemble and bond. The time and expense needed to apply the adhesive and to assemble the stacks is eliminated since they can be counted and bundled within the tooling. The only labor involved will be in fixturing the assembled laminations and in curing and cleaning. The other is a relatively recent approach involving the automatic application of adhesive within the stamping tooling [25], with stacking also taking place within the tooling, negating all assembly labor for the initial stack. Such practices present a potential path to high production of bonded stacks that may provide comparable performance advantages of conventionally bonded cores. Stress relief annealing and the limits to automated bonding Stamped laminations made from thin-gauge electrical steels often benefit from a post stamping annealing process that relieves the stamping-induced stresses created along the sheared edges. The benefits of stress-relief annealing have been well studied; recent work by Johnston [5], and Landgraf and Emura [26], describe these benefits in selected operating regimens. Stress-relief annealing is often called for as lamination thickness is reduced since the percentage of stressed material becomes higher. A limitation to the automated bonding schemes mentioned above is that the laminations will not be able to receive a stress-relief annealing since the adhesive is applied before or during the stamping operation and will not withstand the elevated temperatures of a post-stamping annealing process. An assessment must be made by the machine designer as to the applicability of these methods in light of the potential degradation of magnetic properties resulting from unrelieved stamping-induced stresses.

In-Die Interlocking In-die interlocking, whereby individual laminations are connected inside the stamping tooling by mating protrusions, or tabs, and recesses in the laminations, is commonly used in many high production applications. This process takes advantage of the plastic deformation along the punched edges of stamped metal to provide an interference fit between the tab and the mating recess in the adjacent lamination. This process produces a completely assembled laminated core with no assembly labor. A Brief History The history of interlocking, or staking, of metal parts, has a history almost as long as that of modern metalcraft. All lamination interlocking technology looks back to two patents from the 1890s. First, that of 1890 filed by William Clark, showing a then novel technique for assembling metal sheets [27]. Figure 7 in the schematic at right shows the punching tool A, along with the die plate E, and the group of assembly items B, G and H showing the penetration and locking of the top metal sheet into the bottom sheet. Following shortly was the 1893 patent from James Gould, shown below, describing a method for assembling sheet metal with punched “protuberances.” [28]

The first production of interlocked lamination cores is difficult to ascertain, although the patent history and related literature seems to indicate that direct interlocking of lamination stacks may have been in limited use by the early 1930s [29]. That said, with progressive stamping well established by 1900 [30], the combination of this stamping technology with the Clark and Gould methods and an 1898 patent from Henry Geisenhöner of General Electric, [31] at left, describing a formed lamination feature nesting in a corresponding recess in the adjacent lamination suggests that the manufacturing technology for lamin-ation interlocking may have been developed well before the 1930s.

Though the above paragraphs indicate that the base methodology for the interlocking of electromagnetic cores had been in place for some time, it appears that the first direct description in the patent history is from 1949 for an interlocked transformer core developed by Joseph Sliwiak of Jefferson Electric [32]. He clearly detailed the design and manufacturing technique; not only did he succinctly described the ultimate benefit of his method but gave us the term, interlock, we use today:

It is an object of this invention to provide a core structure and a method of making the same which eliminates the necessity of using additional clamping elements, such as a rivet or a strap, the separate laminations according to my invention being formed so as to interlock with each other. Thus, a less costly structure is provided.

— Joseph Sliwiak U.S. patent 2,671,951

By eliminating secondary assembly methods, the primary benefit of in-die interlocking is immediately clear: there is no assembly labor involved since finished stacks are delivered from the same tooling that stamps the laminations. Just how this process works will also help to explain its possible limitations in producing lamination stacks with thin-gauge of lamination steels. Plastic Deformation … or … The Punch Makes the Hole Metal stamping operations involve the transformation of raw steel into functional pieces through the use of incredible shearing force; these forces are of such magnitude that even the smallest in magnitude are expressed in tons. The impact of a punch striking the sheet stock before driving through the material into the die creates immediate deformation of the material in the region sur-

The sequence at right shows a very simplified and schematized depiction of a

typical press action and the resulting effect on the dimensions of the part. [33]

rounding the punched feature. This deformation is complex in nature and subject to wide variations depending on the mechanical properties of the steel, tool design and manufacture and also tool maintenance. In the finished part, this deformation results in the dimension of that portion of the material that has been punched through the die, “the slug,” having the approximate dimension of the die section, and the portion remaining in the raw material, the “hole,” generally measuring the dimension of the punch. The collapse of material around the punch gives rise to the commonplace though not entirely accurate phrase “the punch makes the hole.” [34] Since the punch and die are separated by a distance equivalent to about 5% of the material thickness (that is, the “die clearance” required to satisfactorily punch the material without undue material deformation or damage to the tooling), one can see that the slug punched through tooling will be about 10% larger in overall dimension, or twice the 5% clearance, than the remaining hole. Interlocking technology takes great advantage of this deformation. As shown in the figure at right [35], the interlocking mechanism is one of stamping a rectangular tab, or other small feature, partially through the material without completely severing it from the base steel material, and then in a second operation pressing the upper tab into the recess of the tab in the adjacent lamination. Interlock sections are often shown diagrammatically as successive layers of steel precisely sheared into a perfectly slip-fit square section, very much as shown in Figure 7 of the Sliwiak patent above, a reconstruction of which is shown at right. [33] In actuality, because the edges of the tabs have experienced the plastic deformation described above they are ever so slightly larger than the recesses below them, and are forced into the recesses creating an interference fit that locks them in place. The diagram at right adapts the prior figure to show the edge deformations that make for the interlocking interference. Thin-gauge lamination steel can produce only narrow interference surfaces, compounding the challenge of interlocking cores made from them. Tooling construction and control regimens, as well as material handling schemes within the tooling, require that the staking punch extend just enough to create the tab or protrusion without severing the feature from the raw material and must also prevent it from not penetrating the material completely. For 0.65 mm (0.025 inch) raw material, this means that in tooling that will see a press stroke of perhaps 75 mm (3 inches), the staking punches must be held in their closing depth, for 0.65 mm (0.025 inch) material, to within 2.0 mm (0.078 inch) below the surface of the steel. That this has been standard practice for over 60 years is a testament to the longstanding expertise and bench skill of tool

designers and toolmakers throughout the industry. This skill is severely tested with thin-gauge materials: for steel that is 0.30 mm (0.012 inch) in thickness, the closing depth of the interlock punch must be held to around 0.90 mm (0.035 inch) below the surface. Thin-gauge electrical steel and the strength of interlocked stacks. The early interlocking schemes were intended for the common lamination thicknesses of their day, in the range of 0.47 to 1.0 mm (0.185 to 0.039 inch). Keep in mind that for these gauges, the die clearance, and the effect of deformation in the dimensions of the mating interlock tabs and recesses, results in a larger surface area as well as greater amount of deformation and the strength of these interlocking joints are substantial. In-die interlocking of these thicknesses develops a strong and unified stack. However, as the thickness of the lamination, and the clearance between punch and die is decreased, both the surface area of the edge of the lamination and the extended deformation decreases, and the strength of the lamination stack is weakened. In a concise exposition of the effects of lamination thickness, coating and interlocking tech-nology on the robustness of an interlocked stack, Nakayama and Kojima showed that an interlocked stack becomes weaker with a reduction in lamination thickness. Their results, summarized in the table at right, show that there is a gradual weakening in interlocking strength in thicknesses down to about 0.36 mm (.014 inch), and then a more dramatic drop in strength as thicknesses continue to drop. The authors continue on to discuss the wide range of fastening values that may be realized by using different interlock shapes and also describe the weakening of the interlock that will be contributed by the surface insulation coatings. [36] Issues Related to Machine Design and Magnetic Performance As with all lamination assembly methods, interlocking can have an important effect on the magnetic properties of the completed interlocked stack. The bare metal of the press-fit tabs and recesses provide a shorting path through the entire stack; Lamprecht, Hömme and Albrecht have shown that interlocking can have a significant impact on eddy current losses, especially as operating frequencies rise. [11] While this study was made of lamination steel of 0.50 mm (0.019 inch), steel that is not properly in the thin-gauge class of materials, other work from Senda, et al [37], and Nakayama and Kojima [36] present similar findings for thinner lamination materials. These papers also discuss the design of the interlock feature, the spacing and number of the interlocks and the impact of such considerations on the magnetic performance of the assembled core. Current Applications and Developments The goal stated by Joseph Sliwiak in his 1949 patent, providing “a less costly structure,” continues to be the driving force for the use of interlocked lamination cores in contemporary electrical machinery. Since the Sliwiak patent, significant research and development has created a sophisticated body of stamping technology capable of interlocking ever thinner laminations. Two main reasons have kept this from being a more common assembly method for use with thin-gauge lamination steels:

Summary Fastened Strength of Interlocked Stacks of Electrical Steel Laminations

Lamination Sheet Thickness Fastened Strength

0.50 mm (0.019 inch) 60 MPa (8.7 ksi) 0.36 mm (0.014 inch) 56 MPa (8.12 ksi) 0.30 mm (0.012 inch) 40 MPa (5.8 ksi) 0.20 mm (0.008 inch) 29.5 MPa (4.28 ksi) Nakayama and Kojima, Interlocking Performances on Non-Oriented Electrical Steels [36]

(conversions approximate)

— The use of thin-gauge electrical steels, until just very recently, has not been typical in high-production scenarios. The tooling required is substantially more costly than production of progressive tooling for individual laminations and requires a substantial production volume to make the additional cost economical. — The technical challenges of interlocking thinner lamination materials described above have proven to be difficult to overcome. That said, interlocking has become achievable in cores with lamination thicknesses of 0.25 mm (0.010 inch) and thicker, with interlocked cores using 0.36 mm (0.014 inch) steel now commonplace.

Thin-gauge silicon steels are primarily specified to reduce eddy current losses in electromagnetic cores and consequently are being designed into a widening array of machines and applications that will experience high excitation frequencies or elevated frequencies resulting from high pole-switching operating regimens. As the production quantities of these specialized motors continue to increase, the demand for greater economy in manufacturing and the resulting lower machine and system costs will drive innovative developments in interlocking technology for thinner laminations. Consumer driven industries where the need for electrical efficiency in the motor or generator is coupled with high production levels are already pushing this technology. As noted in this author’s paper from WMM’14 [2], the Toyota Prius and other automotive propulsion systems are making use of interlocked stacks with lamination in the upper range on thin-gauges thicknesses, 0.25 to 0.30 mm (0.010 to 0.012 inch).

Stamped and interlocked rotor core for 2010

Toyota Prius traction motor, lamination thickness 0.30 mm (0.012 inch) [38]

Prototype and demonstration interlocked stacks have been made from material as thin as 0.20 mm (0.008 inch). and developments in other industries, including medical devices and consumer electronics, are fueling new investigations into the use and assembly of thin-gauge laminations. A recent U.S. patent from Showa, Kurosawa and Tsuge of Minebea presents a claim for computer disk drive spindle motors made from interlocked laminations in thicknesses of 0.10 to 0.20 mm (0.004 to 0.008 inch). [39] The interlock, a diagram of which is shown at left, is significantly weaker than the fastened strength presented by Nakayama and Kojima; for material 0.15 mm (0.006 inch) in thickness the strength is approximately 3 MPa depending on the shape of the interlock. But clearly, this lower strength is not expected to present insurmountable challenges in this particular application.

Advancements both market-driven and technology-driven Thin-gauge electrical steels can offer significant efficiency advantages when used in electrical machines operating at elevated frequencies. Their use continues to be investigated vigorously as new machines are developed that either require or can take advantage of the lower loss profiles these materials can offer. The very fragile nature of laminations made from these materials necessitate a very high level of skill in tooling and fixture manufacture and maintenance and great care in the handling of the laminations. These concerns increase as the thickness decreases. Despite these intrinsic challenges, the stamping of thin-gauge electrical steels has become well-established. Production tooling, whether of compound or progressive design, is capable of stamping all of the thicknesses in the thin-gauge range, and is now being investigated for stamping laminations of 0.05 mm (0.002 inch). As prototype and development programs have moved forward into moderate or high levels of production, economic concerns come to bear and manufacturers are looking to automated assembly of laminations to ease these cost burdens. The methods briefly presented in this paper are applicable to a wide range of lamination designs and machine applications. The use of thin-gauge steels is increasing in the construction of contemporary rotor designs of synchronous reluctance, interior permanent magnet and line start AC motors as manufacturers take advantage of the energy efficiency profiles of these materials and also exploit the possibility of thinner webs and flux-weakening bridges. Production levels for such designs are becoming greater, especially in automotive applications, and interlocking of these types of cores has become possible with the increasing expertise of stamping manufacturers (indeed, the Toyota Prius rotor shown above is an interlocked structure for an interior permanent magnet motor). All motor design and manufacturing involves an assessment of performance versus cost issues that must be addressed. In the case of interlocked stacks such as those shown above, positive decisions were made about the cost benefits of an interlocked stack while the possible performance degradations were either discounted or mitigated through the design of the laminations. There may be instances, though, where the lower magnetic performance of the interlocks will be unacceptable, or the material will be too thin to be successfully interlocked and other methods will be employed. These types of concerns apply to each of the assembly methods and must continually be assessed as machine production continues to increase and as new or improved assembly methods are developed and brought to the industry That decisions about appropriate assembly methods for thin-gauge electrical steels can even be made is the result of years of development in stamping and assembly technology by lamination producers and also by the directed research of machine manufacturers into higher system performance and lower manufacturing costs. Producers and users have pursued these programs individually, in collaboration with academic research centers and also in cooperation with each other. This dynamic environment has resulted in the advances presented in this paper, and, one hopes, will continue to drive new technologies for the assembly of laminated stacks made from these most-promising of electromagnetic materials.

Acknowelgement Many thanks to Dr. Juergen Schneider and also to the organizing committee in Italy for their work putting together WMM ’16. The author gratefully acknowledges the support of Mr. Mark Rippy, President of Proto Laminations, Inc., Santa Fe Springs, California USA and the personal support and keen editorial assistance of Mrs. Patricia Sprague, Crestline, California USA. References [1] Sprague, Steve. 2014. Punching and stacking of thin electrical steels, status and

future. The Proceedings of WMM’14, the Sixth International Conference on Magnetism and Metallurgy, Cardiff, Wales: 225.

[2] — … punching of steel becomes more difficult as the laminations become thinner. Steve Constantinides. 2008. Designing with Thin Gauge. Proceedings of the SMMA 2008 Fall Technical Conference, St. Louis, Missouri USA: slide 39

[3] Schneider, L., W. Braun, G. Senn, and M. Brensing. 2006. Thin electrical steels and their applications. Proceedings of WMM’06, the 2nd International Conference on Magnetism and Metallurgy, Freiburg, Germany: 220.

[4] Constantinides, Steve. 2010. Thin gage silicon-iron for motors, generators, transformers and inductors. Proceedings of WMM’10, the 4th International Conference on Magnetism and Metallurgy, Freiburg, Germany.

[5] Johnston, Gwynne. 2010. Processing and Properties of Thin Electrical Steels. The Proceedings of WMM’10, the Fourth International Conference on Magnetism and Metallurgy, Ghent, Belgium.

[6] International Electrotechnical Commission. 1991. IEC 60404-8-8 Specifications for individual materials Section 8 - Specification for thin magnetic steel strip for use at medium frequencies. International Electrotechnical Commission, Geneva, Switzerland.

[7] ASTM International. 2013. A1086-13, standard specification for thin-gauge nonoriented electrical steel fully processed types. ASTM International, West Conshohocken, Pennsylvania, USA.

[8] Schoppa, A., J. Schneider, and C.-D. Wuppermann. 2000. Influence of the manufacturing process on the magnetic properties of non-oriented electrical steels. Journal of Magnetism and Magnetic Materials 215–216:74–78

[9] Arshad, W.M., Thomas Ryckebusch, Freddy Magnussen, Heinz Lendenmann, Juliette Soularde, Bengt Eriksson, and Bo Malmross. 2007. Incorporating lamination processing and component manufacturing in electrical machine design tools. Conference Record of the 2007 IEEE Industry Applications Conference, Forty-second IAS Annual Meeting: 94–102.

[10] Lamprecht, Erik, Martin Hömme, and Thomas Albrecht. 2012. Investigations of eddy current losses in laminated cores due to the impact of various stacking processes. Proceedings of the 2nd International Electric Drives Production Conference (EDPC) Nuremberg, Germany: 61 – 68.

[11] Krings, Andreas, Shafigh Nategh, Oskar Wallmark, and Juliette Soulard. 2012. Influence of the welding process on the magnetic properties of a slot-less permanent magnet synchronous machine stator core. Proceedings of the 2012 XXth International Conference on Electrical Machines (ICEM), Marseille, France: 1333–1338

[12] Lamprecht, Erik, Martin Hömme, and Thomas Albrecht. 2012. Investigations of eddy current losses in laminated cores due to the impact of various stacking processes. Proceedings of the 2nd International Electric Drives Production Conference (EDPC) Nuremberg, Germany:66.

[13] Schade, T., J. Pflomm, D. Shakirov, and J. P. Bergmann. 2014. Electrical steel stacks for traction motors – fundamental investigations of the weldability. Proceedings of the 58th Ilmenau Scientific Colloquium, Ilmenau, Germany.

[14] David Vegelj, David, Boštjan Zajec, Peter Gregorčič, Janez Možina, and Hidria Rotomatika. 2014. Adaptive pulsed-laser welding of electrical laminations. Strojniški vestnik–Journal of Mechanical Engineering 60, no. 2:106 – 114.

[15] Beckley, Phillip. 2002. Electrical steels for rotating machinery. London: Institution of Electrical Engineers: 55.

[16] Beckley, Phillip. 2002. Electrical steels for rotating machinery. London: Institution of Electrical Engineers: 59.

[17] Yamamoto, Ken-ichi, and Shunji Yanase. 2011. Magnetic properties of non-oriented electrical steels under compressive stress normal to their surface. Przegląd Elektrotechniczny (Electrical Review) 87, no. 9B

[18] Miyagi, Daisuke, et al. 2010 Effect of compressive stress in thickness direction on iron losses of nonoriented electrical steel sheet. IEEE Transactions on Magnetics 6, no. 6 (June).

[19] Personal communication with the author. 2016. David Cutler, Premiere Laminations Engineering.

[20] ASTM International. 2015. A340-15, standard terminology of symbols and definitions relating to magnetic testing. ASTM International, West Conshohocken, Pennsylvania, USA.

[21] ASTM International. 2013. A719/A719M-02 (reapproved 2007), standard test method for lamination factor of magnetic materials. ASTM International, West Conshohocken, Pennsylvania, USA.

[22] International Electrotechnical Commission. 1995. IEC 60404-13 Magnetic materials – Part 13: Methods of measurement of resistivity, density and stacking factor of electrical steel strip and sheet. International Electrotechnical Commission, Geneva, Switzerland.

[23] 3M Corporation. 2011. ScotchWeld™ Epoxy Adhesive/Coating 2290 Technical Data. 3M Corporation, July 2011: 2

[24] National Electrical Manufacturers Association. 2009. NEMA Standards Publication MG-1-2009, Motors and Generators. National Electrical Manufacturers Association, §1.66.

[25] Kielnle & Speiss. 2015. Glulock® and Glulock HT. On-line document: http://www.kienle-spiess.de/glulock-and-glulock-ht.html

[26] Landgraf, F.J.G., and M. Emura. Losses and permeability improvement by stress relieving fully processed electrical steels with previous small deformations. Journal of Magnetism and Magnetic Materials vol 242–245:152–156.

[27] Clark, William H., and William J. Clark. 1890. Method of uniting the edges of sheet metal. U.S. Patent 430,000, filed February 8, 1890, and issued June 10, 1890.

[28] Gould, James, Jr. 1893. Method of uniting metallic sheets. U.S. Patent 512,021, filed August 11, 1893, and issued January 2, 1894.

[29] Phelps, Alva W. and Norman L. Penn. 1929. Assembling machine. U.S. Patent 1,817,462, filed November 29, 1929, and issued August 1931.

[30] Lucas, J.L. 1897. Dies and Die-Making. Providence, Rhode Island, USA: Journal of Commerce:

[31] Geisenhöner, Henry. 1989. Dynamo-electric machine. U.S. Patent 642,599, filed November 7, 1898, and issued February 6, 1900.

[32] Sliwiak, Joseph. 1949. Transformer core and method of making same. U.S. Patent 2,671,951, filed October 3, 1949, and issued March 16, 1954.

[33] Schematic diagram prepared by Steve Sprague. Elemental deformation based on similar diagrams from Andy Spence-Parson. 2008. The formula for successful punching. On-line document: http://www.thefabricator.com/article/punching/the-formula-for-successful-punching

[34] Hendrick, Art. 2006. Die basics 101: part X. on-line document: http://www.thefabricator.com/article/stamping/die-basics-101-part-x

[35] Figure adapted from Gašparin, Lovrenc, and Rastko Fišer. 2011. Influence of asymmetries in stator back iron of PMS motors to the level of cogging torque components. Przegląd Elektrotechniczny (Electrical Review) 87, (March)

[36] Nakayama, Tasei, and Hiroshi Kojima. 2007. Interlocking performances on non-oriented electrical steels. Journal of Materials Engineering and Performance 16, no. 1(February): 7 – 11.

[37] Senda, Kunihiro, Hiroaki Toda, and Masaki Kawano. 2015. Influence of interlocking on core magnetic properties. IEEEJ Journal of Industry Applications 4, no. 4:496 – 502.

[38] Burress, T.A. 2011. Evaluation of the 2010 Toyota Prius hybrid synergy drive system. United States Department of Energy, Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA, page 44, figure 2.63.

[39] Showa, Hideaki, Ryoichi Kurosawa, and Hironobu Tsuge. 2015. Spindle motor and hard disk drive therewith. U.S. Patent 8,941,947, filed October 11, 2013, and issued January 27, 2015.

About the Author and Proto Laminations, Inc. Steve Sprague is editor of the Lamination Steels CD-ROM series from EMERF–the Electric Motor Education and Research Foundation and is a past member of the Board of Directors of EMERF. He is a member of ASTM International’s Committee A06 on Magnetic Properties and is Chair of that group’s Subcommittee A06.02 on Material Specifications and Subcommittee A06.94, the TAG to IEC TC68. In addition, he serves as Technical Advisor on the U.S. National Committee Technical Advisory Group to IEC

TC68 on Magnetic Alloys and Steels. He has given presentations at meetings of the International Conference on Magnetism and Metallurgy, IEEE Industry Applications Society and International Conference on Electrical Machines, the Soft Magnetic Materials Conference, the Incremental Motion Control Systems Society and the Technical and Management Conferences of SMMA–the Motor and Motion Association. After receiving Bachelor of Arts and Master of Fine Arts degrees from the University of California, Irvine, Mr. Sprague has been active in the metal working industry with over 30 years of involvement in the manufacture of electrical laminations for motors, generators and other electromagnetic devices. His ongoing interest is in the development and use of complete and consistent material property data for electrical steels. Proto Laminations, Inc., incorporated in 1989, is located in Santa Fe Springs, California, an industrial community in the Los Angeles metropolitan area, occupying a manufacturing facility of approximately 11,000 square feet (1020 square meters) with 11 employees. We are specialists in custom-made stamped and laser cut electrical laminations for rotating machines supporting the research and development, prototype evaluation and limited production needs of motor and generator manufacturers and academic and industrial research centers worldwide.