I left money on the table when I filed this patent.

Why I know better now.

The Patent I Filed—And the Thirty Years of Lessons It Taught Me

How a Single Composite Bolt Patent Became a Case Study in Leaving Money on the Table

Herbert Roberts, PE • Inventor’s Mind • inventorsmindblog.com

The Bolt That Started Everything

In October 1992 I filed a patent application that would become US 5,292,215 A: Metal Bolt with a Composite Core for Enhancement. The invention was straightforward in concept—a metal bolt shell with a composite fiber core bonded inside it, the fibers arranged in longitudinal and spiral groups to increase tensile strength and reduce thermal axial growth. I was working in aviation R&D, deep in the world of gas turbine engines, and we were exploring how to fabricate metal matrix composites. The problem I was solving was specific: bolted flanges in high-temperature operating conditions lose preload as the metal fastener expands. A composite core with a negative or near-zero coefficient of thermal expansion could counterbalance that growth and keep the joint tight.

The patent issued in March 1994. It has since been cited 27 times by other inventors—companies ranging from Boeing and Lockheed Martin to Rolls-Royce, Siemens, Ford, Samsung Heavy Industries, and an Israeli orthopedic implant company called Carbofix. Those 27 forward citations are a map. They trace the territory I could have claimed but did not, because I was solving one contradiction when the underlying principle solved at least six.

This article is not about regret. It is about methodology. If I had known then what I teach now—the systematic contradiction analysis that TRIZ provides—I would have filed a fundamentally different application. What follows is a forensic reconstruction of what happened, what could have happened, and what every inventor and patent attorney can learn from the gap between the two.

What I Actually Claimed

The five claims of US 5,292,215 define a single embodiment: a metal matrix bolt comprising a metal shell (head, shank, threaded section) with a longitudinal space substantially therethrough, and a composite core of selectively oriented fibers bonded within that space. The fibers are organized into at least one group of longitudinal fibers and at least one group of spiral fibers, the spiral group wrapped clockwise and inclined at 30° to 45° from the bolt axis.

The stated objects of the invention were three: (a) to provide a metal matrix bolt, (b) to provide one with increased tensile strength, and (c) to provide one with reduced axial growth during thermal expansion. The specification describes the benefit of the composite core counterbalancing thermal expansion of the metal shell, thereby reducing preload loss.

What the claims did not specify: any particular matrix material class, any particular fiber material, any fastener form other than a bolt with head-shank-threads, any application domain, or any property modification beyond tensile strength and CTE. The claims were simultaneously too specific in form (metal shell, bolt geometry, spiral fiber architecture) and too silent on the broader principle that made the invention valuable.

The Contradiction I Solved

Every invention worth patenting resolves a contradiction. In TRIZ methodology, we distinguish between technical contradictions (improving one parameter degrades another) and physical contradictions (a system element must simultaneously possess opposite properties). My bolt addressed a physical contradiction:

The fastener must expand thermally with the flange material to maintain compatibility at the joint interface, BUT must not lose preload as temperature rises, which requires dimensional stability the base metal cannot provide.

The resolution was separation in structure: the outer shell expands with the flange while the inner composite core—with its negative or near-zero CTE fibers—counterbalances that growth. The net axial expansion of the fastener is reduced, preload is maintained, and the joint stays tight through thermal cycling. It was an elegant solution to a real operational problem.

But it was one contradiction. The principle of placing selectively oriented reinforcing fibers inside a structural fastener shell could resolve many others.

The 27 Citations: A Map of Unclaimed Territory

Forward citations are not just a metric of patent influence. They are a forensic record of what the market needed that your patent pointed toward but did not capture. Each citation represents an inventor who read your disclosure, recognized something useful in it, and then filed claims in adjacent territory you left open. When you read them carefully, they cluster into exactly the categories a systematic contradiction analysis would have revealed before filing.

Cluster 1: Weak Matrix Reinforcement

The largest cluster of citations came from inventors who took my core insight—reinforcing fibers inside a structural fastener shell—and applied it to matrix materials that need the fiber core for basic structural adequacy, not merely for CTE control. The Carbofix orthopedic patents are the most striking example: at least six citations covering bone screws, bone implants, self-tapping inserts, and multi-layer composite bone screws. These are fasteners where the shell is a biocompatible polymer or composite, inherently too weak to carry surgical loads alone. The fiber core is the primary load path. Samsung Heavy Industries filed on fiber-reinforced plastic bolts for marine applications. Boeing’s hybrid fastener patents (US 9,238,339 and US 9,623,612) bridged both worlds—composite cores in fasteners designed for composite airframe assembly where galvanic compatibility and weight matter more than thermal expansion.

The contradiction I missed: The fastener must be made of a lightweight, biocompatible, or non-conductive material (polymer, ceramic, wood) BUT must carry structural loads comparable to metal fasteners. My fiber-core architecture solves this directly, but my claims required a metal shell.

Cluster 2: Brittle Material Toughening

Siemens filed EP 2,497,963—a screw or bolt comprising two different materials—addressing the contradiction that a fastener must be hard enough to resist wear and thread damage but tough enough to absorb impact without catastrophic fracture. A ductile fiber core inside a brittle shell provides crack arrest and energy absorption: the fastener may crack but it does not shatter. The reverse configuration—a brittle high-strength core inside a ductile shell that holds fragments together—is equally valid.

The contradiction I missed: The fastener must be hard enough to resist wear, corrosion, or thread damage BUT must be tough enough to absorb impact and avoid catastrophic fracture. My dual-material architecture resolves this through separation in structure, identical to my CTE solution.

Cluster 3: Controlled Failure

One of the earliest citations was US 5,567,096—a shear-pin system for a logging truck bunk, filed just two years after my patent issued. The inventor used a composite-core fastener as a designed failure point. The most recent citation, Airbus Operations’ US 2022/0170499 (Structural Fuse), does the same thing three decades later. In both cases the core-shell interface or fiber architecture is engineered to create a known-energy failure mode.

The contradiction I missed: The fastener must carry load under normal conditions BUT must fail predictably at a defined threshold to protect the larger structure. The fiber architecture and core-shell bond provide precise control over failure energy.

Cluster 4: Vibration Damping

Ford’s US 10,808,747 (Frictionally Damped Fasteners, 2020) used the core not for strength or CTE management but for vibratory energy dissipation. A viscoelastic or friction-generating core inside a metallic shell absorbs vibration that would otherwise fatigue the joint or loosen the fastener.

The contradiction I missed: The fastener must be rigid enough to maintain clamp load BUT must dissipate vibratory energy to prevent fatigue or loosening.

Cluster 5: Compression Members and Non-Bolt Forms

Rolls-Royce’s composite spacer patent (US 7,897,241) took my reinforced-core-inside-a-shell concept and applied it to a compression member with no threads at all. GE’s dynamic load reduction system explored yet another structural variant. The Boeing thermoplastic compression molding patents extended the manufacturing methodology into entirely new form factors.

What I missed: My claims specified head-shank-threads—the classic bolt geometry. A broader filing could have encompassed studs, pins, dowels, rivets, spacers, standoffs, bushings, inserts, anchors, clamps, U-bolts, nails, and staples—any structural member where a reinforced core within a shell resolves a property contradiction.

Six Contradictions, One Principle

The following table summarizes the contradictions that a systematic pre-filing analysis would have identified. I solved Contradiction 1. The market—through 27 forward citations over three decades—solved the other five using my principle.

# The Fastener Must… BUT Must Also… Resolution via Core-Shell Architecture

1 Expand thermally with flange material Maintain preload as temperature rises Negative-CTE fiber core counterbalances metal shell expansion (MY PATENT)

2 Be made of lightweight / biocompatible / non-conductive material Carry structural loads comparable to metal Fiber core provides primary load path inside weak-matrix shell

3 Resist wear, corrosion, and thread damage (hardness) Absorb impact without catastrophic fracture (toughness) Ductile core arrests cracks in brittle shell, or brittle core stiffens ductile shell

4 Carry load under normal conditions Fail predictably at a defined threshold Engineered core-shell interface creates known-energy failure mode

5 Be rigid enough to maintain clamp load Dissipate vibratory energy Viscoelastic or friction-generating core absorbs vibration within rigid shell

6 Provide mechanical clamping Manage electrical or thermal pathway Core fiber selection controls conductivity independent of structural role

The Claim That Could Have Been

Had I abstracted the principle before drafting claims, the independent claim might have read something like this:

A composite fastener comprising: an outer shell of a first material; and a core of selectively oriented reinforcing elements bonded within said shell; wherein said core modifies at least one mechanical, thermal, or physical property of said fastener relative to a fastener consisting solely of said first material.

The dependent claims would then cascade through four dimensions of specificity:

Shell material classes: metals, alloys, polymers, ceramics, glass, wood, biocompatible materials, and combinations thereof.

Fastener forms: bolts, studs, screws, pins, dowels, rivets, spacers, standoffs, bushings, inserts, anchors, clamps, nails, and staples.

Reinforcing element architectures: longitudinal, spiral, helical, woven, braided, random, and layered configurations.

Property modifications: coefficient of thermal expansion, tensile strength, shear strength, fracture toughness, vibration damping, electrical conductivity, thermal conductivity, and controlled failure load.

This structure would have placed me upstream of most of those 27 forward citations. The engineering was already done. The abstraction was not.

The Non-Technical Lesson: Organizational Isolation

There is a second lesson embedded in these citations that has nothing to do with claim drafting. Several of the forward-citing inventors worked at the same company I did. The 1997 composite fastener patent for high-temperature environments (US 6,045,310) was filed by the same corporate entity—five years after mine—addressing adjacent territory in the same operational domain. We were not building on each other’s foundations. We were working in parallel, in separate pockets, unaware of each other’s filings.

This is not unusual. In large R&D organizations, patent filing strategy rarely extends beyond the immediate innovation team. The engineer files on the specific problem solved. The patent attorney drafts claims around the specific disclosure. Neither has visibility into what other teams within the same company have filed, are filing, or could file. The result is a portfolio of isolated point solutions rather than a coordinated claim fence around a core principle.

The organizational contradiction mirrors the engineering one: the company needs small, focused teams to generate inventive output (autonomy, speed, deep domain expertise) but also needs broad awareness of its own IP landscape to avoid redundant filings and to build strategic portfolios. The very isolation that enables innovation prevents coordination.

The resolution—as with any TRIZ contradiction—is separation. Separate the generation of inventive concepts (keep it small, keep it focused, keep the domain expertise concentrated) from the strategic positioning of the intellectual property (which requires a portfolio-level view across all teams, all domains, all pending and granted claims). Most companies in the 1990s did not have that second function operating with any real authority. Many still do not.

Takeaways for Patent Attorneys

If you are prosecuting patents for inventors—whether in-house or outside counsel—the story of US 5,292,215 offers several actionable lessons.

1. Push the Inventor Past the First Contradiction

When an inventor walks into your office with a solution, they are almost always presenting the resolution of one specific contradiction in one specific context. Your job is to ask: what other contradictions does this architecture resolve? The inventor solved CTE mismatch in a bolted flange. What about strength deficiency in a polymer fastener? Brittleness in a ceramic fastener? Controlled failure in a structural fuse? Each additional contradiction is a potential dependent claim family—or, better, support for a broader independent claim.

2. Abstract the Principle Before Drafting the Independent Claim

The independent claim should capture the principle, not the embodiment. If the inventor says “metal bolt with composite fiber core,” your claim should say “composite fastener comprising an outer shell of a first material and a core of reinforcing elements.” Every material limitation, every geometric limitation, every application limitation in the independent claim is territory you are giving away. Push those specifics into dependent claims where they add value without narrowing the fence.

3. Inventory the Form Factors

Fasteners are not just bolts. Screws, studs, pins, rivets, spacers, inserts, anchors, clamps, and nails are all compression or tension members that could benefit from a reinforced-core architecture. If the principle works in a bolt, ask whether it works in every other form factor that performs a similar mechanical function. Claim the genus, not just the species.

4. Cross-Reference the Client’s Own Portfolio

Before filing, search your client’s existing patents and pending applications. Are other teams within the same organization working on adjacent problems? Could the current invention serve as a platform patent that supports or broadens existing filings? The attorney who only sees the invention in front of them will draft narrow claims. The attorney who sees the portfolio will draft strategic ones.

5. Read the Forward Citations of Prior Art

This is an underused technique. When you review the prior art your inventor cites or that the examiner raises, look at the forward citations of those references—especially the ones filed after your inventor’s priority date. They show you where the market went next, which means they show you where your claims should extend to preemptively.

Takeaways for Inventors

If you are an inventor—solo, small team, or embedded in a large organization—the lessons are equally direct.

1. Learn to See Contradictions, Not Just Solutions

I solved a CTE contradiction. I did not see the strength contradiction, the toughness contradiction, the controlled-failure contradiction, or the damping contradiction—all of which my architecture could resolve. TRIZ provides a systematic framework for identifying contradictions before you begin solving them. If you only see one contradiction, you will only file one set of claims. The principle almost always reaches further than the first problem that prompted it.

2. Separate the Principle from the Embodiment

My invention was not a metal bolt with composite fibers. My invention was the principle of placing selectively oriented reinforcing elements inside a structural fastener shell to resolve property contradictions between the core and shell. The bolt was one embodiment. The bone screw was another. The structural fuse was another. The frictionally damped automotive fastener was another. I described one. The market found the rest over the next thirty years.

3. Span the Design Space Before Filing

Before you finalize your disclosure, ask yourself three questions. First: what other materials could serve as the shell? Second: what other forms could the fastener take? Third: what other properties could the core modify? Write down every combination that is physically plausible. Each combination is either a dependent claim in your application or a continuation filing that extends your priority date into new territory.

4. Get the Right People in the Room

The person who invented the solution is rarely the person best positioned to see the full scope of the principle. You need someone in the room—a patent strategist, a TRIZ practitioner, a materials generalist, or simply a colleague from a different domain—who can ask the abstraction questions. Not “how does this bolt work?” but “what else could this architecture do?” The person who asks that question before filing is worth more than the 27 citations that answer it afterward.

5. Audit Your Own Work Retrospectively

If you have existing patents, pull them up on Google Patents and read the forward citations. They will tell you, with perfect hindsight, exactly what territory you left unclaimed. That hindsight is not wasted—it trains your eye for the next filing. Every patent you have ever written is a case study in scope. Study them.

The Fracture Surface

In forensic failure analysis, the fracture surface tells you everything. It records the origin of the crack, the direction of propagation, the forces that drove it, and the material properties that either resisted or accelerated failure. You cannot lie to a fracture surface. It is an objective record of what actually happened, independent of what anyone intended or believed.

A patent’s forward citation record is its fracture surface. It tells you where the original scope cracked open, which directions the market propagated through the gaps, and which material—which specific unaddressed contradictions—allowed the breach. Reading it thirty years later, I can see exactly where the stress concentrations were in my original filing. The origin was in the independent claim: too narrow on material, too narrow on form, too narrow on function. The propagation ran through every industry that needed a reinforced fastener for a reason I had not considered.

The money I left on the table was not in the engineering. The engineering was sound. The money was in the abstraction—the step back from the specific to the general, from the embodiment to the principle, from one contradiction to six. That step is what TRIZ provides systematically, and it is what I teach now through the Inventor’s Mind because no inventor should have to wait thirty years and 27 forward citations to learn it.

---

Have you looked back at past work and though “what was I thinking?" Tell us about it. Thank you.

Herbert Roberts, PE, is a Licensed Professional Engineer with 32 years in aviation R&D, 62 patents, and over eight years of forensic engineering consulting for attorneys.

He publishes the Inventor’s Mind at inventorsmindblog.com