An "Equivalent Substitution" That Triggered a Full-Machine Recall

On one project, because the original carbon steel grade was in short supply, an engineer substituted another material following the principle of "same strength class, same corrosion class." The data on paper matched perfectly, and sample testing passed. But six months into mass production, a batch of products began to fracture at customer sites, with failures concentrated at the first thread, and fractures showing typical delayed-fracture characteristics.

The problem was that after substitution, the material's hydrogen embrittlement sensitivity had changed: the new material had higher tensile strength and a coarser grain size, so under the same plating and hydrogen-relief process the residual hydrogen content was higher. On paper "the strength class is the same," but the actual material sensitivity and process window were already different. This shows material substitution is never as simple as "comparing parameter tables"; it involves chain-reaction changes across mechanical properties, heat treatment, surface treatment, and the corrosion system.

The Four Dimensions of Substitution Impact

Breaking the impact of substitution into parts, what is controllable versus what is a red line becomes clear: mechanical properties (tensile strength, yield strength, toughness, fatigue strength—controllable but requiring re-verification); heat treatment (quenching and tempering temperatures, hardness band, hardenability—must be re-established); surface treatment (pretreatment method, coating adhesion, hydrogen-relief parameters—must be re-verified); and the corrosion system (potential difference with mating parts, coating compatibility—most prone to going out of control).

Why is the corrosion system most prone to going out of control? Because corrosion is a system problem, not a part problem. Replacing the bolt with a material of "better" corrosion resistance, if it creates a larger potential difference with the clamped part, will instead accelerate corrosion of the clamped part. A typical example: replacing a carbon steel bolt with a stainless bolt screwed into aluminum—the bolt itself is fine, but the aluminum corrodes much faster.

Red Line 1: Strength Class Must Not Be Lowered on a Hunch

Red line judgment: any substitution that lowers the strength class requires a complete re-verification of the joint.

The effect of lowering the strength class is not linear; it simultaneously changes three things: upper preload drops—the upper preload is proportional to the material yield strength; dropping from grade 10.9 to 8.8 lowers yield strength from 940 to 640 MPa, reducing the upper preload by about 32%; slip resistance drops—friction reserve is proportional to preload, so a 32% preload drop lowers slip resistance by the same 32%; and the fatigue stress amplitude rises relative—while the stress amplitude itself is unchanged, the allowable range corrected by the Haigh diagram changes after the mean stress drops.

Specific judgments: grade 10.9 to 8.8 requires re-verification and is usually not feasible (32% drop in upper preload); grade 8.8 to A2-70 requires re-verification (yield strength drops from 640 to 450 MPa); downgrading to a low-strength class is prohibited as a direct substitution; and shrinking the size at the same time is prohibited (effects stack and cannot be simply estimated).

Red Line 2: Corrosion Protection Must Not Be Judged Only by Salt-Spray Hours

Red line judgment: after substitution, the complete corrosion system formed by "coating + substrate + mating part" must be re-verified.

Three effects that are often missed: potential-difference effect—after changing the substrate or coating, the galvanic relationship with the mating part changes; the potential difference of the original "carbon steel bolt + aluminum" setup may widen several-fold once stainless is used; coating compatibility—some coatings have markedly different adhesion on specific substrates; for example, zinc-aluminum coating works well on carbon steel but may have insufficient adhesion on stainless due to the passivation film; and the area-ratio effect—galvanic corrosion severity relates to the anode-to-cathode area ratio, and a small anode paired with a large cathode corrodes fastest; a small bolt joining a large stainless plate is the most dangerous combination.

Five verification methods: overall corrosion resistance per GB/T 10125 with a neutral salt spray test (NSS), common criteria being no red rust at 480 h / 720 h / 1000 h; cyclic corrosion using CCT, which is closer to real atmospheric conditions than NSS; galvanic corrosion via salt spray or immersion testing of the assembled combination, focusing on whether the mating part corrodes faster; coating adhesion by cross-cut or bend test, with no peeling or flaking; and coating thickness by magnetic gauging or metallography, within drawing limits.

Red Line 3: Hydrogen Embrittlement Risk Must Not Be Evaluated Incidentally

Red line judgment: any substitution that raises tensile strength or changes plating and hydrogen-relief conditions requires re-verification of hydrogen embrittlement.

Hydrogen embrittlement sensitivity is usually divided by tensile strength at 1000 MPa, in three ranges: R_m < 1000 MPa (grade 8.8 and below) carries low risk, and conventional hydrogen relief suffices; R_m ≥ 1000 MPa (grades 10.9, 12.9) enters the sensitive range, and the hydrogen-relief process must be strictly controlled and verified; R_m > 1200 MPa (grade 12.9 and above) is high risk, and hydrogen-free processes (zinc-aluminum coating, mechanical plating) should be preferred.

There are three traps in substitution: strength "just over the line"—the original material at 990 MPa and the substitute at 1050 MPa, though both in the same class range, cross the 1000 MPa sensitive line, markedly raising hydrogen embrittlement risk; grain-size change—the coarser the grains, the stronger the tendency for hydrogen to segregate at grain boundaries; and hydrogen-relief process not adjusted in step—different materials have different hydrogen diffusion coefficients, so the original relief time may be insufficient.

Verification Method

Execute per GB/T 3098.17-2000 (Pre-load test for the detection of hydrogen embrittlement): assemble to a specified pre-load (typically about 90% of yield strength), hold at a specified temperature (e.g., 80°C) for 48 h, then check for fracture or cracking.

A Decision Table for Allowed vs. Prohibited Substitution

Classify by risk: low risk—substituting a grade within the same class and material system (e.g., 35CrMo to 42CrMo) or changing suppliers without changing the process, requiring verification of mechanical properties and hardness band, redo of salt spray and hydrogen embrittlement sampling, full inspection of the first batch plus a small trial assembly; medium risk—changing the surface-treatment system (electro-zinc plating to zinc-aluminum coating), requiring redo of corrosion verification and torque-clamping force tests (friction coefficients change); high risk—changing the property class requires redo of joint verification, friction-coefficient testing, and tightening-spec adjustments; changing the material system (carbon steel to stainless/aluminum/titanium) additionally requires galvanic corrosion and thermal-expansion-difference checks; very high risk—changing both size and class, treated as a new product requiring redesign and revalidation. For high-safety-class joints (braking, steering, load-bearing lifting points, battery pack fastening), any substitution must go through the complete design-change process.

Conclusion: Ask Four Questions Before Substituting

Material substitution is not impossible, but it must not be done "casually." Before approving any substitution, you should be able to answer four questions clearly: strength and preload—does the joint calculation still hold after substitution, and which items need re-verification? corrosion system—has the corrosion relationship with mating parts changed, and was combined-part corrosion verification done? hydrogen embrittlement risk—does tensile strength exceed 1000 MPa, does the hydrogen-relief process need adjustment, and was a pre-load hold test done? assembly process—has the friction coefficient changed, does tightening torque need adjustment, and can on-site tools and cycle time cope? Only when all four questions have evidence-based answers is the substitution safe. Conversely, if the reason for substitution is merely "a bit cheaper" or "a bit faster lead time," the risk it brings will almost certainly exceed the cost saved—in fasteners, the cost of one failure is often the cost of an entire machine or an entire batch.