US 5809638

I'll search for authoritative information on patent 5809638, including any litigation activity.

Let me verify litigation status and the patent's family/current owner with additional targeted searches.

US Patent 5,809,638 — Verification Summary

Note on method: I verified this record against the full authoritative patent text (Google Patents mirror of the USPTO grant, fetched 2026-09-30) and via live web searches. I did not find any Federal Circuit (CAFC) litigation or 2026 docket activity involving this patent number. I'll flag the limits of that search below.

Bibliographic Data (authoritative)

Family: A division, US 6,237,214 B1 ("Apparatus for manufacturing laminated parts with center interlock," filed 1998-08-05), and a continuation, US 6,984,913 B2 ("Lamination stack with center interlock," filed 2000-12-18), share the 1992-10-26 priority. Both also expired.

Abstract (verbatim)

"The present invention involves a method and apparatus for forming a centrally interlocked stack of laminations for rotors. A centrally located, circular indentation is formed in each lamination. The indentation provides corresponding projections and depressions in the laminas which may be interlocked by engaging adjacent depressions and projections. Using a progressive die assembly, laminas are blanked from stock sheet material and punched to form the projection/depression. The machine for manufacturing the laminations includes a blanking station including a punch for forming the projections/depressions, and a rotating and stacking station. The projection/depression forming station includes a counterforce which is applied when punching the projection/depression so that the stock material is relatively flat. The laminations are blanked, then rotated, and finally stacked together to form a lamination stack."

Plain-Language Overview of the Independent Claims

The core inventive concept: instead of traditional circumferential interlock tabs (which force rotational alignment to multiples of 360°/number-of-tabs and are hard to place on small-diameter rotors), the laminations are locked together by a single large circular indentation at the center of each lamina. Because the interlock is rotationally symmetric, adjacent laminas can be rotated by any angle before interlocking.

Claim 1 — Basic two-lamina method (method of making a rotor lamination stack):

1. Form a first lamination in strip stock with (a) circumferentially spaced conductor-slot openings and (b) a centrally located, generally circular indentation that creates a depression on one face and a projection on the other.
2. Blank the lamination out of the stock.
3. Stack it onto a second lamination that also has a central circular indentation — during which the first lamination is automatically rotated relative to the second by an "infinitely adjustable" angle to set the desired conductor-slot skew.
4. After rotation, interlock the two indentations so the projection of one nests into the depression of the other.

Claim 8 — Multi-lamina production method:
Same concept scaled to a full stack: form many laminations each with the central circular indentation, blank them, and stack them — automatically rotating each lamination relative to the accumulating stack by an infinitely adjustable angle to achieve the desired skew, then interlocking adjacent indentations after each rotation.

Claim 16 — Two-lamina method with thickness compensation (n = number of conductor slots):
Same steps as claim 1, but the rotation step additionally requires rotating the first lamination relative to the second by about 360/n° to spread out stock-thickness variations (this is now a required element of the independent claim, not a dependent add-on).

Claim 18 — Multi-lamina method with thickness compensation:
Same steps as claim 8, but the stacking step must additionally rotate at least one lamination relative to the stack by about 360/n° for stock-thickness compensation. (Note: the granted claim contains an apparent typo — "thickness variations is said stock material" — which I quote literally rather than correct.)

Notable Dependent Claims

• Claims 2, 10 — the forming punch is driven to a depth of about half the stock thickness.
• Claims 3, 11 — a counterforce of at least ~10% of the blanking pressure is applied during indentation-forming to flatten the indentation surface (specification states a preferred range of 10–20%; the counterforce may be backed by a die spring or nitrogen cylinder).
• Claims 4, 12 — a knockout punch directly presses the indentation to force the projection into the adjacent depression.
• Claims 5, 6, 13, 14, 17, 19 — thickness compensation via an additional rotation of ~360/n° or a multiple of 360/n°.
• Claims 7, 15 — the rotation is accomplished by rotating a choke barrel that holds the stack.
• Claim 9 — herringbone skew: rotate laminations in one direction for a first portion and the other direction for a second portion.

Litigation / CAFC 2026 Docket Status

• I found no record of any 2026 CAFC appeal or docket entry involving US 5,809,638. Given the patent expired in 2010 and the maintenance-fee lapse was recorded in 2010, active appellate litigation is unlikely.
• The only L.H. Carbide enforcement matter surfaced in search was an older case, L.H. Carbide Corp. v. The Piece Maker Co. — which concerns L.H. Carbide's lamination-tooling patent portfolio generally; I could not tie it to the '638 patent specifically.
• Uncertainty flag: My tools were web searches (Google Patents, public litigation aggregators, law-firm CAFC digests) rather than a direct authenticated query of USPTO Patent Center / PACER / the CAFC docket. I therefore cannot certify that no 2026 proceeding exists — only that none was found. Anyone needing a definitive litigation check should query PACER and the CAFC docket directly for this patent number.

Also worth noting: the specification's figures and text reference the assignee's earlier patents (U.S. 4,619,028; 4,738,020; 5,087,849; 5,123,155) as incorporated prior art, and the cited prior art includes US 4,728,842 (Carbet) and US 5,075,150 (Linton & Hirst) for projection/depression interlocks.

I'll research litigation involving US 5,809,638 across patent litigation sources.

Let me search more targeted litigation databases.

Let me dig into specific sources, including CourtListener and the L.H. Carbide cases.

Bottom line

I found no litigation specifically asserting U.S. Patent 5,809,638 in any of the sources I was able to reach. I want to be explicit that this is a negative finding based on limited searches, not a certified "no litigation exists" conclusion — see the caveats at the end.

What I searched

• General web searches combining the numbers "5809638" / "5,809,638" with litigation terms (infringement, lawsuit, complaint, court).
• Targeted searches against Unified Patents and CourtListener-oriented content.
• Assignee-based searches ("L.H. Carbide" patent infringement action).

Important caveat on the search results: keyword searches on the bare number returned substantial noise. Specifically, hits referencing a "'580 patent" (e.g., eSpeed, Inc. v. BrokerTec USA, L.L.C., the trading-system patent, and Lucent Techs. v. Gateway, 580 F.3d 1301) and FloodBreak v. AMI concern completely different patents/citations, not 5,809,638. Per your instruction, I am not treating those as matches.

The one related enforcement action I located (not an assertion of '638)

Why this cannot be '638 litigation: US 5,809,638 issued on 1998-09-22 (application 07/966,876, filed 1992-10-26). A Northern District of Indiana decision in 1994 necessarily predates issuance of the '638 patent, so the '638 patent could not have been asserted in that case. The suit was presumably premised on other L.H. Carbide patents in the same family line (compare the '638 specification's incorporation by reference of U.S. Pat. Nos. 4,619,028; 4,738,020; 5,087,849; and 5,123,155).

I am reporting it only because it is an L.H. Carbide patent-enforcement case against a lamination-tooling competitor and is the closest thing to a "hit" the searchers surfaced — not as litigation involving 5,809,638.

Context that bears on the plausibility of any '638 suit

From the patent record itself:

• Issuance: 1998-09-22. Enforceable window: roughly 1998-09-22 to 2010-09-22.
• Termination: The patent expired 2010-09-22 due to nonpayment of maintenance fees under 37 C.F.R. § 1.362 (lapse recorded 2010-09-22; fee lapse notice 2006-03-22 for the 8th-year payment, then a reminder in 2010). So any infringement suit on the '638 patent must have been filed between 1998 and the 2010 lapse.
• Family: One divisional — US 09/129,146, issued as US 6,237,214 B1 ("Apparatus for manufacturing laminated parts with center interlock") — and one continuation — US 09/739,174, issued as US 6,984,913 B2 ("Lamination stack with center interlock").
• No litigation is listed on the Google Patents page for US 5,809,638 (it lists 15 prior-art citations, 11 third-party citations, and a family of three U.S. applications, but no litigation section — consistent with no reported suits).

Recommended next steps (sources I could not fully execute against)

To convert this from a negative search result to a verified one, these should be run directly:

1. Unified Patents litigation portal — https://portal.unifiedpatents.com/litigation/caselist (search U.S. 5,809,638 and the family members 6,237,214 and 6,984,913).
2. CourtListener / RECAP — docket and opinion search on "5,809,638" and on plaintiff "L.H. Carbide."
3. Docket Navigator or Lex Machina — patent-specific docket coverage back to the late 1990s (free search engines generally do not cover 1990s–2000s district court dockets well).
4. PACER for N.D. Ind. and E.D. Mich. (the Piece Maker transfer destination), and for the Northern District of Indiana/Indiana state courts where L.H. Carbide was/is based.

Confidence statement

• High confidence: No litigation asserting US 5,809,638 appears in the publicly indexed sources I could reach; the patent expired in 2010 for fee nonpayment.
• Low confidence / explicitly unknown: Whether a pre-2010 district court suit on the '638 patent exists in a docket-only (non-indexed) record. I did not find the Piece Maker docket number or filing date, and I found no case number, filing date, or outcome for any suit naming 5,809,638 as an asserted patent. I will not fabricate those identifiers.

If you want, I can help you draft the exact query strings for Docket Navigator/Lex Machina and the Unified Patents and CourtListener endpoints to close this out.

I'll verify the structured data against what's publicly searchable before writing anything up.

Let me run a couple more targeted checks before concluding.

Proceedings overview

There are zero AIA trial proceedings on file for US 5,809,638. The structured "PTAB proceedings on file" block reports no IPR, PGR, or CBM from the USPTO Open Data Portal, and my independent web checks surfaced no petition or Board decision naming this patent. My searches did return several hits referencing a "'580 patent" (e.g., Rembrandt Wireless Technologies' US 8,023,580 and its IPR2020-00037/IPR2014-00518 family, and a Bell Semiconductor/NXP dispute) — those are different patents with an unrelated docket number coincidence; none of them is US 5,809,638. Do not let a search engine's "'580" alias pull those proceedings into this file.

The breakdown is therefore: 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials. Nothing has been tested at the Board, which means the claims of the '638 patent are untested, not "hardened." The more important fact for a defendant is not the PTAB silence at all — it's the patent's legal status: the '638 patent expired on 2010-09-22 for failure to pay maintenance fees (37 CFR 1.362), well before any AIA-trial window would have been useful. A patent that lapsed in 2010 is not an IPR target; it's a § 286 damages-bar argument.

No proceedings to enumerate. With a count of zero, the per-proceeding template (proceeding number, panel, institution decision, FWD, settlement, appeal) has nothing to populate. I will not manufacture proceeding numbers to fill the sections.

What I verified, and the limits of that verification:

• The structured ODP block (canonical source per the task instructions) reports no AIA trials: US5809638A on Google Patents — the "Cited By"/"Citations" tables list only patents and printed publications, no PTAB entries.
• I ran four targeted web searches (US5809638 IPR PTAB, "5809638" patent PTAB L.H. Carbide, "5,809,638" IPR petition, "US 5,809,638" / "5809638" reexamination litigation). Nothing responsive to this patent appeared.
• I did not have direct query access to PTAB E2E (https://ptacts.uspto.gov/ptacts/) or CourtListener in this session, so treat this as a high-confidence-but-not-exhaustive null. The structured ODP data is the authoritative statement, and it agrees.

Strategic summary

Claim status. Because no IPR, PGR, or CBM ever reached a Final Written Decision, no claim of US 5,809,638 has been canceled, disclaimed, or held unpatentable by the Board. Claims 1–19 (the patent issued with 19 claims; 1, 8, and 16 are the independent method claims, with 9 offering herringbone skew, 5/6/13/14/17/19 offering the 360/n° thickness-compensation rotation, and 2/3/10/11 offering the punch-depth and counterforce limitations) are all "untested" — not "sustained." That distinction matters: an untested claim carries no estoppel shield, but it also carries no Board-endorsed validity that a defendant must overcome. Before you build any defense on the merits of these claims, note the prior-art record that was cited: US 4,728,842 (Carbet) and US 5,075,150 (Linton & Hirst) both disclose projections/depressions interlocking adjacent laminations, and US 4,110,895 / US 4,160,182 (Mitsui) disclose the progressive-die stacking art — a § 103 combination attack on the "centrally located circular indentation" claims would have been the obvious route had anyone filed.

Estoppel landscape. Section 315(e)(2) estoppel is a null set here — no petitioner exists, so no party (and no privy of any party) is barred from raising any §§ 102/103 ground in district court or the ITC. Every prior-art ground remains fully available to a defendant. That is the one genuine upside of the Board's silence.

Pattern signals. There are none to read. No repeat petitioner, no serial-filer dynamic, no Unified Patents-style defensive aggregator in the chain, and no patent-owner appeal activity — because there was never a proceeding. The patent's ownership history is simple: filed 1992-10-26 by inventor Thomas R. Neuenschwander, assigned to L.H. Carbide Corporation (recorded 1993-01-11, effective 1992-10-19), with the small-entity status dropped in 2001. The family has three members, all now lapsed: US 5,809,638 (this patent), US 6,237,214 B1 (divisional, apparatus — "Apparatus for manufacturing laminated parts with center interlock"), and US 6,984,913 B2 (continuation filed 2000-12-18 — "Lamination stack with center interlock"). If a demand letter cites any member of this family, the same expiration analysis applies and should be run member-by-member.

The signal that actually matters for a defendant. The absence of PTAB activity is ordinarily a weak negative signal, but here it is explained by something dispositive: the '638 patent's maintenance-fee clock stopped. It was granted 1998-09-22, and the file shows the 12-year maintenance fee was never paid — "2010-04-26 Maintenance fee reminder mailed," "2010-09-22 Lapse for failure to pay maintenance fees," and "2010-10-18 PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362." The anticipated term under 35 U.S.C. § 154(c) (17 years from grant, which exceeds 20 years from the 1992 filing) would have run to 2015-09-22, but the patent did not live that long. A lapsed patent cannot be revived in a way that supports new infringement accrual, and no AIA trial can be filed against conduct occurring after the lapse.

Recommended next steps

1. Do not treat this as an IPR matter — treat it as a standing/damages matter. The threshold defense is the 2010-09-22 lapse. Confirm the lapse against the USPTO Patent Center maintenance-fee record for US 5,809,638 and against the fee record for US 6,237,214 B1 and US 6,984,913 B2 — the Google Patents family table shows all three as "Expired - Fee Related," but verify each independently before you send a letter.
2. If a demand letter cites the '638 patent, run the § 286 math immediately. No recovery is available for infringement more than six years before the complaint is filed. With the patent's enforceable life ending in 2010, a complaint filed today reaches back to, at earliest, 2020 — a period during which the patent was already expired. On its face, that leaves an empty damages window absent some revival theory, which you should force the plaintiff to articulate in writing before engaging on the merits.
3. Preserve the merits defense anyway. Because no claim has ever been canceled or even construed by the Board, you get no free claim constructions and no institution-stage findings. If the plaintiff has a revival argument you cannot defeat, the merits fight is a fresh § 102/§ 103 case, and the Carbet ('842), Linton & Hirst ('150), and Mitsui ('895/'182) references cited during prosecution are your starting point rather than anything the Board produced.
4. Correct any adverse research before you rely on it. As noted above, the "'580 patent" references that appear in IPR2020-00037 and the Rembrandt Wireless family are to US 8,023,580 — an unrelated wireless-modulation patent. If opposing counsel or a vendor's docket search hands you a proceedings list under a "'580" alias, confirm the number digit-for-digit before citing it. For any genuinely relevant Board document, go to PTAB E2E and for any Federal Circuit disposition to CourtListener — but expect both to return nothing for this patent.
5. Flag the null as provisional. The ODP ingest is the canonical source and it says zero, and my searches agree. But I could not query PTAB E2E myself this session. Before filing anything that asserts "no PTAB proceedings exist," re-run the check against E2E by application number (07/966,876) and patent number, and re-confirm on the 2010-09-22 lapse.

I'll verify the assignment record and owner history with targeted searches before writing up.

Ownership & Assignment Analysis — US 5,809,638

Data-fidelity note up front: I could not open an authenticated session on USPTO Assignment Center (assignmentcenter.uspto.gov) or the legacy Assignment Search UI; my tooling is web search. The assignment facts below are taken from the USPTO assignment data as mirrored in Google Patents' legal-events table for this patent (fetched 2026-09-30), plus corroborating searches. Where a field (notably the correspondent of record) was not exposed in that mirror, I say so rather than guess. Anyone needing a certified reel/frame pull should query Assignment Center directly.

Inventors

Unusual-pattern check — negative. The classic fire-sale precursor (all inventors bolting the assignee within 12 months of filing) is not present here. Neuenschwander is a career L.H. Carbide tooling inventor, appearing on the assignee's patents continuously for roughly three decades — from the 1980s–90s core portfolio (US 5,087,849; 5,123,155; 5,349,741 — the patents this specification incorporates by reference) through US 6,000,119; 6,131,268; 6,163,949; 6,195,875; 6,636,137; 6,745,458; 7,062,841, and still as inventor on US 2014/0139067 A1, assigned to L.H. Carbide Corporation. That is a stable, single-employer inventor track, not an abandoned one.

Original assignee

L.H. Carbide Corporation (Google Patents renders the assignee as "LH Carbide Corp"), Fort Wayne, Indiana — the entity named in the 1993 recorded assignment and listed as original assignee on the face of the patent.

• Business: contract tool-and-die maker founded in 1966 by Leon Habegger; core competency is progressive stamping dies and precision stamped metal products. Its sister division LH Stamping specifically produces interlocked electrical steel cores and loose laminations — i.e., the very article this patent claims a method of making.
• Did they ship a product embodying the claims? Yes. The claimed method is a progressive-die lamination-interlocking process, and the assignee's business is building exactly those dies and running them to make interlocked rotor/stator cores. This is an operating-company patent, not an idle asset.
• Current status: Operating, private, no bankruptcy and no third-party acquisition found. The business is now organized under a holding company, LH Industries, with four divisions (LH Carbide, LH Stamping, LH Medical, LH Controls), still headquartered at 4420 Clubview Dr., Fort Wayne, IN 46804. Habegger is Chairman; family members run the operating units. Sources: Craft.co profile ("LH Industries — formerly LH Carbide … Status: Active"); Junior Achievement Northern Indiana profile of Leon Habegger ("Leon Habegger developed … VeriLok … LH Carbide holds more than 40 United States and worldwide patents"); FWBusiness ("LH Industries … is now comprised of four divisions").
• Caveat: the corporate recorded owner of this patent was never changed to "LH Industries." The LH Carbide → LH Industries transition appears to be a trade/organizational identity rather than a recorded chain-of-title event, and I found no USPTO change-of-name or merger record for it. If a formal re-recordation exists it did not surface in the Google Patents legal-events mirror.

Assignment timeline

There is exactly one recorded assignment against this patent, and zero post-issuance assignments.

• 1992-10-19 (executed) / recorded 1993-01-11 — Reel 006371 / Frame 0476
  • Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
  • Assignor: Thomas R. Neuenschwander
  • Assignee: L.H. Carbide Corporation (Indiana)
  • Correspondent: Not exposed in the sources I could retrieve. The Google Patents assignment event for reel 006371/0476 lists only owner name, assignor, reel/frame and effective date (1992-10-19); no recording attorney or firm is shown. I will not invent one. This is the single most valuable field in an NPE-chain analysis and it is the one field I could not verify — flagging explicitly as a gap.
  • Context: Routine employee-to-employer assignment of rights, executed one week pre-filing. Not a fire-sale, not a securitization, not a transfer to an asserter.

No other assignment records exist for this patent. All remaining post-issuance legal events are fee/administrative only and should not be mistaken for title changes:

The 2001 small-entity withdrawal is a common false-positive in troll screening; here it is unremarkable — it reflects the assignee's revenue growth past SBA thresholds, not a change of owner, and no assignment accompanies it.

Family note (not verified in Assignment Center): the division US 6,237,214 B1 (filed 1998-08-05) and continuation US 6,984,913 B2 (filed 2000-12-18) share the 1992-10-26 priority. Google Patents shows no family-level assignment events for them either, but I did not independently pull their reel/frame records — treat their chains as unverified.

Timeline diagram

timeline
    title Ownership of US 5809638
    1992 : Inventor executes assignment
         : Filed 26 Oct 1992
    January 1993 : Recorded reel 006371 frame 0476
    1998 : Patent issued 22 Sep
    2001 : Small entity status dropped
    2010 : Lapsed for unpaid maintenance fees

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. No assignment to any "IP / Holdings / Ventures / Licensing" entity exists. The only recorded assignee is the operating tool-and-die manufacturer itself (reel 006371/0476, 1992-10-19).

2. Known asserter in the chain — NOT PRESENT. Neither the original assignee nor any successor matches a public NPE roll (Acacia, Marathon, IV, IPNav, Wi-LAN/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Spangenberg entities). The chain terminates at L.H. Carbide Corporation / LH Industries, a Fort Wayne manufacturer.

3. Repeat correspondent across the chain — NOT ASSESSABLE. With only one record and that record's correspondent field unexposed in the source I reached, the recurrence test cannot be run. This is a data gap, not a clean bill of health. Note separately that a single appearance would not be a finding under this rubric anyway.

4. Cascading transfers — NOT PRESENT. Zero consecutive transfers of this patent. No chained LLCs, no shared correspondent addresses (none exist to share).

5. Pre-litigation transfer — NOT PRESENT. No assignment within 6 months of any infringement suit naming this patent; there is no post-issuance assignment at all. The nearest litigation artifact is L.H. Carbide Corp. v. Piece Maker Co., 852 F. Supp. 1425 (N.D. Ind. 1994) (venue/personal-jurisdiction ruling; transferred to E.D. Mich. under 28 U.S.C. § 1406(a)) — but that case predates this patent's 1998-09-22 issuance by four years, so it cannot be a suit on the '638 patent. It almost certainly concerned the assignee's earlier incorporated-by-reference patents.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 for L.H. Carbide Corporation or LH Industries found in search; the company is described as active with ~$21M revenue and 64+ employees (Owler).

7. Privateering — NOT PRESENT. No operating company → NPE transfer on record, so no arm's-length assertion vehicle was created.

8. Defensive aggregator — NOT PRESENT (but inverse not triggered). The chain does not terminate at RPX, AST, LOT, Unified, or OIN. It terminates at the original operating company, which simply let the patent lapse in 2010.

Verdict

Insufficient data — where "insufficient data" is defined by this rubric as no records, or only the original assignment.

Justification: the '638 patent has exactly one recorded assignment, reel 006371/0476, executed 1992-10-19 and recorded 1993-01-11, moving title from inventor Neuenschwander to L.H. Carbide Corporation, and it was never reassigned again before lapsing for non-payment of maintenance fees on 2010-09-22. That is not enough chain to support any NPE verdict, and the affirmative evidence that does exist runs the other direction — the current/last owner is a Fort Wayne progressive-die manufacturer (now LH Industries, divisions LH Carbide and LH Stamping) that builds dies and interlocked lamination cores embodying the claims, with a single long-tenured inventor and no asserter anywhere in the chain. I deliberately do not select "Operating-company assertion," because that verdict requires the owner to be suing actual competitors on this patent, and no suit on the '638 patent was found.

Verification links:

• USPTO Assignment Search (query 5809638): https://assignment.uspto.gov/patent/index.html
• Assignment Center: https://assignmentcenter.uspto.gov/
• Google Patents legal events for this patent: https://patents.google.com/patent/US5809638/en

Open items a definitive check should close: (a) the correspondent of record on reel 006371/0476, which I could not retrieve; (b) assignment records for family members US 6,237,214 and US 6,984,913; (c) whether any non-recorded (unrecorded) transfer occurred — unrecorded assignments would not appear in Assignment Center but would affect standing.

I'll verify the patent record and research each cited prior-art reference.

Let me research the key interlock-related prior-art references in more detail.

Prior-Art Analysis — US 5,809,638

Patent under analysis: US 5,809,638 A, "Method for manufacturing laminated parts with center interlock" — App. No. 07/966,876, filed/priority 1992-10-26, granted 1998-09-22, inventor Thomas R. Neuenschwander, original assignee L.H. Carbide Corporation (Google Patents records "LH Carbide Corp"). 19 claims, independent claims 1, 8, 16, 18.

Scope note / search limits: I verified the record against the authoritative full text and via live web search. Google Patents is a mirror of the USPTO grant; the fetched text is treated as authoritative per my operating rules. I was able to pull full text for only some of the cited references (notably US 5,075,150, US 4,728,842, US 4,264,663). Descriptions for the remaining references are drawn from their titles plus the characterizations of them that appear inside the cited references' own "prior art" discussions (e.g., the Carbet and Linton & Hirst patents characterize US 2,975,312, US 4,538,345, US 4,364,169, US 3,203,077, US 3,202,851). I flag where confidence is lower. I did not complete a full-text pull of every reference; a definitive § 102 analysis would require the full specifications and claims of each.

1. Legal framework applied

This patent has a 1992 priority date, so pre-AIA 35 U.S.C. § 102 governs. All 15 examiner-cited references published before 1992-10-26 by more than one year, so each is available as § 102(a)/(b) art (early ones also § 102(e) if a US application's pre-issue date matters).

Anticipation under § 102 requires a single reference disclosing every limitation of the claim, arranged as in the claim. Anything short of that is, at most, § 103 obviousness. This distinction matters here: the cited corpus is a history of interlock geometries (tabs, teeth, projections/recesses) and of stacking/rotating machinery, but I found no single reference that discloses all elements of independent claims 1, 8, 16 or 18. The strongest references are near-misses discussed in § 4.

The four independent claims require, in combination:

• (A) circumferentially spaced conductor-slot openings;
• (B) a centrally located, generally circular indentation creating a depression on one face and a projection on the other;
• (C) blanking the lamination from stock;
• (D) automatically rotating the lamination relative to the stack by an infinitely adjustable angle (claims 1, 8) or an angle sufficient for the desired skew plus about 360/n° (claims 16, 18) — i.e., rotation for skew and thickness compensation;
• (E) interlocking the indentations after rotation so a projection enters an adjacent depression.

Element (D) is the crux, and it is the element most of the cited art does not disclose in combination with (B).

2. The 15 examiner-cited references

Note: the patent's own specification also lists, as incorporated by reference, U.S. Pat. Nos. 4,619,028; 4,738,020; 5,087,849 and 5,123,155 ("all assigned to the assignee"). 4,738,020 = item 13; the latter three correspond to US 4,619,028, US 5,087,849 and US 5,123,155, which appear in the family-cited list (see § 5) but not in the examiner's 15. I flag that 4,619,028 in particular is a separate L.H. Carbide document and is not part of the 15-reference citation set.

Per-reference assessment

1. US 2,671,951 — Jefferson Electric (1954). Foundational transformer-core lamination assembly and method. Antedates any interlock-tab or projection/depression teaching relevant here. Potential § 102 relevance: essentially none as to claims 1/8/16/18 — no central circular indentation, no rotation-for-skew. Background art only (§ 102(a)/(b) but not anticipating).

2. US 2,975,312 — Globe Union (1961). Per the later Carbet patent's own discussion, this reference "discloses the use of interlocking projections and recesses having the same identical outlines and dimensions." That is, it teaches the projection-into-recess interlock concept underlying element (E). Potential § 102 relevance: bears on the interlocking-indentation step of claims 1, 8 (and arguably claims 16, 18) and the lamination-as-article aspects, but it does not disclose a centrally located, generally circular indentation (element B) or automatic rotation by an infinitely adjustable angle for skew (element D). No complete anticipation; § 103 material.

3. US 3,202,851 and 4. US 3,203,077 — General Motors (1965). Aligning/fastening laminations of a dynamoelectric machine and a "fastening assembly and procedure." These are early tab/mechanical-fastening approaches and alignment during stacking. Potential § 102 relevance: they touch the stacking-and-fastening steps of claims 1 and 8 (and the aligning concept that underlies claims 7/15's barrel rotation), but not the central circular indentation or the rotation-to-skew. Not anticipating; § 103/background.

5. US 4,112,734 — L. Schuler GmbH (1979). "Aligning and stacking arrangement" — i.e., machinery that receives laminations and stacks them in alignment. Potential § 102 relevance: bears on the stacking step and the rotating/stacking station of claims 1, 8, and on claims 7/15 (a barrel that holds the stack), but it is an alignment/stacking apparatus, not a central-indentation interlock with infinitely adjustable skew rotation. Not anticipating.

6. US 4,160,182 and 7. US 4,110,895 — Mitsui Mfg. (1978/1979). "Laminated core manufacture" and "Apparatus for manufacturing laminated cores" — progressive-die core-making equipment. Potential § 102 relevance: general apparatus context for claims 1/8 (progressive-die manufacture and stacking); no teaching of the centrally located circular indentation or the rotation-for-skew limitation. Not anticipating.

8. US 4,264,663 — Licentia (1981). Verified full text: a laminated pack of closely superposed sheets held by mechanical crimping/pinching, where teeth on each sheet are bent down into recesses of the next sheet at crimping locations (typically corners), formed in a combined stamping-and-stacking tool. Potential § 102 relevance: directly relevant to the projection-into-recess interlock of claims 1/8 and to the in-die forming-and-stacking concept; but the deformations are peripheral corner teeth, not a central circular indentation, and there is no rotation-for-skew or thickness-compensation teaching. Not anticipating; strong § 103 backdrop for the "interlock by deformed material" concept.

9. US 4,364,169 — Nippondenso (1982). Method of producing a stator iron core (cited in the Carbet patent among "various other uses for the projections and recesses"). Potential § 102 relevance: general interlock/projection-recess background for claims 1/8; not a central circular indentation, no rotation-for-skew. Not anticipating.

10. JP S58-66557 A — Hitachi (1983). Japanese published application, manufacturing method of an electric motor core. Potential § 102(b) printed publication relevance only at a general level; no evident disclosure of elements B or D. Not anticipating (subject to full-text verification).

11. US 4,538,345 — Siemens (1985). Manufacture of a pack of coated metal laminas (also cited by Carbet as projection/recess prior art). Potential § 102 relevance: general laminated-pack construction; not the central-indentation interlock or skew rotation. Not anticipating.

12. US 4,728,842 — Carbet / Benson D. Martin (1988). Verified full text. A multistation progressive die forms, simultaneously, a recess in one face and a projection on the other of a lamina; the projection carries exterior longitudinal ridges that produce a compressive interference fit into the recess of the adjacent lamina. Its divisional US 4,809,429 claims the die. Critically, this reference is a method-and-apparatus progressive-die interlock patent for rotors/stators — the closest structural analogue to the '638 interlock-forming step. Potential § 102 relevance: its "punch forms a recess while a lower die forms a projection" step is close to element (E) and to the die-assembly forming step of claims 1/8; but (i) the interlock is not a centrally located, generally circular indentation, (ii) it is a compressive interference fit rather than the press-flattened central indentation, and (iii) it discloses no automatic rotation of a lamination relative to the stack by an infinitely adjustable angle for skew, and no 360/n° thickness compensation. Not a complete anticipation of claims 1/8/16/18; it is the leading § 103 reference for the interlock geometry and the die station.

13. US 4,738,020 — L.H. Carbide (1988). Same assignee as the '638 patent; "Method for manufacture of laminated parts." This is one of the four L.H. Carbide patents the '638 specification expressly incorporates (along with US 4,619,028, US 5,087,849, US 5,123,155). These disclose the barrel/choke stacking, rotation of the stack before each lamination is blanked, and thickness-variation compensation by rotation that the '638 specification identifies as the prior-art system it improves upon. Potential § 102 relevance: the single most material prior art on the rotation/thickness-compensation method steps (elements D of claims 1, 8, 16, 18) — though the prior-art rotation there is 180° (or a minimum of 360°/number-of-tabs), which the '638 specification itself distinguishes, and it teaches tab interlocks rather than a central circular indentation. This reference makes the "rotate for thickness compensation" concept (§ 102/§ 103 against claims 16 and 18) squarely old; the '638 contribution is the infinitely adjustable / less-than-360/n minimum rotation and the central circular indentation interlock that makes it possible.

14. GB 2,206,453 A — Linton & Hirst Ltd (1989). "Laminated cores." Filed the same day (1987-06-22) and by the same company as US 5,075,150 (item 15); this appears to be the GB member of the same Linton & Hirst lamination-interlock family (see item 15). Potential § 102 relevance: same subject matter as item 15 — projections/depressions interlocking laminations. My search of the INPI record for the related Linton & Hirst EP application (EP 0 353 029, priority GB 8,817,875 of 1988-07-27) confirms the family concerns "packs of laminations … interfitting formations … deformed to engage … to clamp … together." Treated together with item 15; not independently anticipating the '638 claims for the same reasons.

15. US 5,075,150 A — Linton and Hirst (1991). Verified full text. The most on-point reference. It discloses a pack of laminations in which each lamination is connected to its adjacent lamination by a single connection provided by a projection that projects into a depression, the projection and depression being aligned on a common axis perpendicular to the lamination (i.e., a central, single interlock), produced by a projection punch cooperating with a die that extrudes metal past a radiused/conical lip. The pack is expressly "torsionally flexible." This is the closest disclosure of elements (B) and (E): a centrally located single projection/depression interlock. Potential § 102 relevance: potentially anticipatory of the central-indentation interlock structure recited in claims 1 and 8 to the extent those claims are read structurally, and highly material under § 103 for the "rotate the laminations to any angle" advantage (torsional flexibility ≈ adjacent laminations not rotationally locked). It is not a complete anticipation of independent claims 1/8/16/18, because it does not disclose (i) forming conductor slots and a lamination stack for a rotor, or (ii) the automatically rotating … by an infinitely adjustable angle sufficient to define a slot having the desired skew angle step (element D) that the '638 claims require as a process step. Note also the '638 specification distinguishes projection/depression depth: it drives the punch to "about half the thickness" (claims 2, 10) whereas Linton & Hirst is directed to depths greater than 50%.

3. Bottom line on § 102 anticipation

No single cited reference anticipates independent claims 1, 8, 16, or 18. The novelty resides in the combination of (a) a centrally located, generally circular indentation interlock that is rotationally symmetric so that any relative angle is permissible, and (b) automatically rotating the lamination by an infinitely adjustable angle to set skew (and by 360/n° or a multiple thereof for thickness compensation). The closest art — Linton & Hirst's central single projection/depression connection (US 5,075,150 / GB 2,206,453) and L.H. Carbide's own rotation-compensation method (US 4,738,020) — would frame the strongest § 103 challenge, individually or in combination, rather than a § 102 anticipation. Dependent claims 2/10 (punch to ~half thickness) and 3/11 (counterforce ≥ ~10% of blanking pressure) find no counterpart in the cited art at all and appear the most defensible.

4. Secondary citations on the record (not in the examiner's 15)

These appear in the file as "Family Cites Families" — references cited during prosecution of the family members or the applicant's earlier patents — and are relevant because the '638 specification expressly incorporates some of them:

• L.H. Carbide rotation/stacking patents: US 5,087,849; US 5,123,155; US 4,619,028 (all incorporated by reference in the '638 specification; see the internal note in § 2). These are the "aforementioned U.S. Patents" whose 180°-rotation tab-interlock system the '638 patent sets out to improve — highly material § 102/§ 103 art against the rotation aspects of claims 1, 8, 16, 18.
• Interlock geometry: US 4,614,780 (GE), US 4,951,377 (GE), US 4,979,285 (Martin), US 5,174,009 (Martin, "Apparatus for forming a stack of interlocked laminations"), GB 2,226,459 B (Linton & Hirst), IT 1,286,450 B1 (Corrada), DE 2,344,193 B2, DE 3,227,681 A1, US 3,590,208 (Licentia), US 4,897,916 (Coils, Inc.), JP H06-14481 (Mitsubishi).
• Alignment/stacking: US 5,142,178 (Emerson Electric), US 5,349,741 (L.H. Carbide).
• Later (post-'638 priority, citing the '638 family): US 5,641,349 (Greenway); US 5,894,182; US 5,992,003; US 6,002,191; US 6,018,207; US 6,223,417 — these post-date the '638 priority date and therefore are not prior art to it; they are forward citations only.

5. Caveats

1. Full-text verification is incomplete. I obtained and read the full text of US 5,075,150, US 4,728,842, and US 4,264,663 and the '638 patent itself. For items 1–11, 13, 14 I relied on titles, the patents' own cross-characterizations, and partial web retrieval. A rigorous § 102 chart should be built from the full specifications and claims of each reference, which I did not exhaustively retrieve.
2. GB 2,206,453 A ↔ US 5,075,150 are treated as likely members of the same Linton & Hirst family based on identical filing date (1987-06-22) and assignee; I did not confirm the exact family link (the EP search report for the related Linton & Hirst EP 0 353 029 lists complementary-lamination subject matter). Treat them as related but verify if this matters.
3. "Potentially anticipates" above is used in the strict sense of could a single reference disclose every claim limitation. My conclusion is that none of the cited references does so for the independent claims; several are strong § 103 references.
4. My web searches hit a step limit before I could retrieve every reference's full text; where I say "subject to full-text verification," that is a real gap, not a satisfied check.

I'll ground this obviousness analysis in the actual content of the cited prior art. Let me research the key references.

Let me pull the substance of the remaining key references — the Linton & Hirst pack, the L.H. Carbide rotation/indexing patents, and the older projection/recess references.

Obviousness Analysis — US 5,809,638 under 35 U.S.C. § 103

Patent: US 5,809,638 A, Method for manufacturing laminated parts with center interlock (Neuenschwander; L.H. Carbide Corp.), filed 1992‑10‑26, granted 1998‑09‑22, 19 claims (independents 1, 8, 16, 18), expired 2010.

Governing law note (important, because of the 1992 filing date): This is a pre‑AIA patent. Obviousness is judged under pre‑AIA § 103(a) as construed by Graham v. John Deere and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007). Because I am analyzing this in 2026, KSR's "expansive and flexible" standard applies, and it materially strengthens the combinations below. Statutory‑bar dates below use the 1992‑10‑26 filing as the § 102(b) cutoff of 1991‑10‑26.

1. Level of ordinary skill in the art (POSITA)

A person of ordinary skill would have a degree in mechanical engineering (or equivalent tool‑and‑die trade education) plus several years of hands‑on experience in the design and operation of progressive die assemblies for dynamoelectric‑machine laminations — i.e., someone familiar with: (a) blanking/stacking stations and choke barrels; (b) interlock tab, projection/recess and "dimple" interlocks; (c) rotational indexing of laminas to set skew; and (d) strip‑thickness variation and its effect on stack parallelism. Both asserted references are squarely within this person's ordinary working knowledge.

2. The critical claim‑construction points

The "infinitely adjustable angle" phrase is a functional result that inherently follows from a circular, rotationally symmetric interlock. Under § 103, a functional recitation of an inherent result of a known structure adds no patentable weight.

3. The prior‑art landscape (from the '638 record)

Applicant's own admission. The '638 specification itself characterizes the Carbide family as a known system in which "the barrel holding the lamination stack is automatically rotated before each lamina … to compensate for variations in thickness while still properly skewing the conductor slots," and expressly states that projection/depression interlocks and interlock tabs were "well known." Under § 103 these admissions may be used as evidence of what the art taught. See the '638 "Description of the Related Art."

4. The controlling combination

Combination A — Claims 1 and 8

Linton & Hirst '150 (or its § 102(b) sibling GB 2,206,453) in view of L.H. Carbide '020/'849/'155, optionally with Carbet '842.

Mapping:

Motivation to combine (articulable, and more than hindsight):

1. Same field of endeavor, same problem. Both references address progressive‑die manufacture of interlocked dynamoelectric‑machine lamination stacks with skewed conductor slots.
2. The secondary reference solves the express shortcoming of the primary's interlock. The Carbide rotation patents admit (and the '638 spec repeats the admission) that interlock tabs constrain rotation to discrete increments (90°/180°; at minimum 360°÷number‑of‑tabs) and that this costs cycle time. Linton & Hirst '150 was directed precisely at enabling continuous, any‑angle relative rotation of the laminas ("torsionally flexible contact"). A POSITA seeking arbitrary skew angles (including the herringbone and zero‑skew cases the '638 recites) would look to exactly this reference to remove the discrete‑increment constraint. This is a textbook KSR "simple substitution of one known element for another to obtain its predictable result," and "use of a known technique to improve a similar device in the same way."
3. Reasonable expectation of success. Substituting a circular central interlock for a tab interlock preserves the known function (holding the laminas together) while adding the known benefit (unrestricted indexing). No new mechanism is required; Carbet '842 confirms that forming a projection in a lower die and a recess in the punch face is routine in the same progressive die that performs the stacking.
4. "Obvious to try." Where there is a finite number of identified, predictable interlock geometries and an express need for arbitrary skew, the KSR "obvious to try" rationale independently supports the combination.

Result: Claims 1 and 8 would have been obvious over Linton & Hirst '150 + Carbide '020/'849/'155 (further in view of Carbet '842).

Combination B — Dependent method claims

Thus, on a claim‑by‑claim basis, claims 1, 2, 4–8, 10, 12–19 are strong § 103 candidates; claims 3, 11 (and possibly 9) are the only meaningful battleground.

5. Counter‑arguments the patent owner would raise (and my assessment)

(a) Linton & Hirst teaches away — "torsionally flexible." L&H '150 deliberately leaves the projection shorter than the depression (its cl. 2) so the pack remains twistable after assembly; the '638, by contrast, presses the indentation (claim 4) to seat a tight, rotation‑fixing interlock that preserves skew. Assessment — weak. Claim 1 does not require a twist‑locking or interference fit; it requires only that "said projection … extends into said depression." The twist‑fixing feature lives only in dependent claims 3/4/11/12, and Ploran '312 and Carbet '842 independently teach forcing the projection into the recess for a holding, non‑twisting fit. KSR warns that "a patentee simply cannot … assume that a person of skill in the art will be led … only by the problem the patentee was trying to solve." The combination (L&H circular geometry + Ploran/Carbet tight seating + Carbide indexing) is directly motivated by the desire for a rigid, skewed rotor core.

(b) The counterforce limitation (claims 3, 11). No reference in the record expressly states a counter‑force measured as a percentage of blanking pressure applied during indentation forming. This is the single most defensible limitation. But: (i) Ploran '312 already applies a backing/punch pressure to the recesses while forming the joint; (ii) Carbet '842's entire stated purpose is reducing deformation/distortion of the plates — the identical concern the '638 counterforce addresses; (iii) applying controlled die‑cushion/"counter‑punch" back‑pressure to flatten a formed feature and to avoid fracture is elementary sheet‑metal forming; and (iv) the numeric "≥ 10 %" is a result‑effective amount, and the '638 spec concedes the value "is dependent on the actual materials used and the surface area," which undercuts any criticality argument. Under KSR, reciting a routine, result‑effective process parameter obtained by ordinary experimentation does not save the claim. Assessment: likely obvious, but this is where a patentee would focus.

(c) Herringbone skew (claim 9). The '638 spec asserts herringbone was impractical with tab interlocks (it "would require an additional processing station"). I did not locate a reference in the '638 record that expressly teaches forming a herringbone by reversing the indexing direction between stack halves. A POSITA desiring the known herringbone geometry could readily reverse rotation direction, and L&H '150 permits rotation either way, so the claim is probably obvious to try — but I flag this as the least‑grounded mapping in the chart.

(d) Objective indicia / secondary considerations. I found no evidence of commercial success with an established nexus, no copying, no licensing attributable to this claim set, and no litigation that would supply such evidence (the patent expired in 2010). The specification does identify a "long‑felt need" for interlocks on small‑diameter rotors and for herringbone/straight skew; that establishes a problem, not nonobviousness of a solution assembled from known elements (KSR; In re Kahn). Absent nexus evidence, secondary considerations do not outweigh the primary Graham factors.

(e) A note on KSR vs. the 1998 allowance. The '638 issued in 1998, i.e., under pre‑KSR practice, where examiners often required an express teaching, suggestion or motivation to combine. The very reference that most squarely discloses the central, any‑angle circular interlock (Linton & Hirst) is listed on the face of the patent, yet the examiner allowed the claims — most likely because no reference expressly combined that interlock with programmed skew/thickness‑compensation indexing. Under KSR, that gap is filled by the predictable‑use/obvious‑to‑try rationale, so the validity picture today is materially weaker than it was in 1998. (This is an analytical observation about the § 103 standard, not a legal conclusion.)

6. Bottom line

Overall: For all independent claims and most dependents, a prima facie § 103 case exists, and the strongest single reference (Linton & Hirst's torsionally flexible single circular projection/depression) is already of record in the '638 file. The genuine points of attack for a patent owner are (i) the counterforce‑to‑blanking‑pressure limitation of claims 3 and 11, and (ii) the absence of an express herringbone teaching for claim 9.

7. Verification limits / uncertainty flags (per my operating rules)

• I verified the text of claims and key passages of US 5,075,150, US 2,975,312, US 4,728,842, US 5,123,155, and the '638 record via live retrieval (Google Patents, Justia, FreePatentsOnline, EveryPatent). URLs: patents.google.com/patent/US5075150, patents.google.com/patent/US4728842, patents.google.com/patent/US2975312, patents.google.com/patent/US5123155, patents.google.com/patent/US5809638.
• I did not retrieve the full text of US 4,979,285 (Benson/Martin), US 5,174,009 (Martin), US 4,264,663 (Licentia), or US 4,951,377 (GE). Based on titles/abstracts they concern connecting/interlocking stacks and core‑distortion control (potentially additional § 103 ammunition, especially for claims 3/11), but I cannot state their disclosures with confidence. Anyone relying on this analysis should pull those texts.
• I could not confirm the exact station‑by‑station disclosure of US 4,738,020 individually; I relied on the Carbide family's collective disclosure of automatic skew indexing and thickness‑compensation rotation, which the '638 specification itself admits.
• US 5,075,150 issued 1991‑12‑24 (less than one year before the '638 filing), so it is § 102(a)/(e) art only if not also available via GB 2,206,453 A (1989‑01‑05), which is § 102(b) art and is on the '638 face. The substantive disclosure is thus unambiguously prior art either way.
• Claim 18 of the '638 contains an apparent typographical error ("thickness variations is said stock material"). Per my rule, I quote it literally and do not correct it; it does not affect the substance of the § 103 analysis.