In the development and testing phase of a diesel engine, we encountered a recurring issue where multiple crankshafts made from nodular cast iron experienced torsional fractures at the keyway located on the front small end. This crankshaft, manufactured from grade QT700-2 nodular cast iron, undergoes a standard process sequence: casting, normalizing, machining, fillet rolling, and final cleaning and packaging. The failure occurred specifically at the keyway that accommodates a flat key, connecting to the timing gear, with the outer journal linked to a pulley. As engineers involved in this investigation, I will detail our comprehensive approach to diagnosing the root cause and implementing a structural improvement, emphasizing the material behavior of nodular cast iron throughout.
The primary objective was to determine whether the failure originated from material defects, manufacturing imperfections, or design-related stress concentrations. We employed a multi-faceted analytical methodology. Chemical composition was assessed using an ARL EasySpark 1160 optical emission spectrometer. Mechanical properties were evaluated via tensile testing on an MTS C43 universal testing machine, adhering to GB/T 228.1-2010. Microstructural examination was conducted with an Olympus GX71 metallographic microscope, following the rating standards of GB/T 9441-2009. Finally, to understand the stress distribution, we performed finite element analysis (FEA) using the ABAQUS software suite. The consistent use of nodular cast iron in these components made it crucial to verify its conformance to specifications.
Our initial step involved a macroscopic examination of the fracture surfaces. The fracture appearance was reminiscent of a ratchet pattern, indicating multiple initiation sites. The crack origins were distributed around the circumference of the small-end journal, with the final rupture occurring on the side opposite the keyway. Closer inspection revealed that the primary crack initiation site was precisely at the contact point between the journal surface and the keyway. The journal surface exhibited significant circumferential torsional wear, and the crack propagation direction was approximately 45 degrees relative to the shaft axis.
We then proceeded with microstructural analysis by sectioning samples from the fractured keyway area, focusing on the crack origin, keyway root, sidewall, and bottom. At the crack origin, the nodular graphite exhibited a spheroidization grade of 2 and a nodule size of 5, which are acceptable for QT700-2. After etching, the matrix structure was primarily pearlitic, rated at approximately 95% pearlite, meeting technical requirements. However, at the root of the keyway, we observed locally deformed graphite and poor roundness of the root profile. The keyway sidewalls showed no obvious defects. Intriguingly, several micro-cracks were found at the keyway bottom, and the matrix in that area contained about 1% carbides. The microstructural integrity of the nodular cast iron at the critical crack origin site was, therefore, not the direct culprit.
Mechanical property testing was conducted on specimens extracted from the crankarm of the failed crankshaft. The results are summarized in the table below, confirming that the material’s tensile strength, yield strength, elongation, and hardness all conformed to the specifications for QT700-2 nodular cast iron.
| Test Item | Tensile Strength Rm (MPa) | Yield Strength RP0.2 (MPa) | Elongation δ (%) | Hardness (HBW) |
|---|---|---|---|---|
| Technical Requirement | ≥ 700 | ≥ 400 | ≥ 2 | 241–302 |
| Measured Value 1 | 845 | 566 | 4.0 | 285 |
| Measured Value 2 | 833 | 535 | 4.1 | 283 |
| Measured Value 3 | 794 | 608 | 4.3 | 278 |
To rule out batch-to-batch variation, we traced the chemical composition of the molten iron from the same production lot as the failed crankshaft. The data, presented in the following table, confirms that all elements were within the specified range for QT700-2 nodular cast iron. The consistency in composition further underscored that the failure was not due to a fundamental material flaw in the nodular cast iron.
| Element | C | Si | Mn | P | S | Cr | Mo | Ni | Cu | Mg |
|---|---|---|---|---|---|---|---|---|---|---|
| Specification Range | 3.08–3.30 | 2.49–3.13 | 0.55–0.61 | 0.02–0.042 | 0.01–0.022 | 0.022–0.032 | ≤ 0.01 | ≤ 0.010 | 0.8–1.1 | 0.022–0.040 |
| Sample 1 | 3.12 | 2.53 | 0.58 | 0.021 | 0.013 | 0.027 | 0.003 | 0.006 | 0.993 | 0.029 |
| Sample 2 | 3.20 | 2.56 | 0.56 | 0.022 | 0.011 | 0.026 | 0.006 | 0.004 | 0.982 | 0.032 |
| Sample 3 | 3.18 | 2.57 | 0.57 | 0.021 | 0.011 | 0.026 | 0.006 | 0.004 | 0.97 | 0.036 |
| Sample 4 | 3.16 | 2.77 | 0.58 | 0.021 | 0.120 | 0.029 | 0.006 | 0.004 | 0.997 | 0.032 |
Based on these findings, we concluded that the nodular cast iron material itself—in terms of composition, microstructure, and mechanical properties—was not defective. The observed deformed graphite at the keyway root was attributed to tooling pressure during machining, and the micro-cracks at the bottom were likely secondary features formed during torsional overload. Nevertheless, to address any potential machining influence, we modified the machining parameters for the keyway. The milling speed was reduced, and a root radius of R0.3 was explicitly machined to improve geometry and eliminate graphite deformation. Subsequent testing of crankshafts with this improved machining still resulted in fracture at the same location—the contact point between the keyway edge and the journal surface. This persistent failure pointed decisively towards a design-induced stress concentration issue.
The geometry of the flat key keyway became the focus. The keyway length was 22 mm on a journal length of 25 mm, meaning it occupied almost the entire engaged length. During operation, stresses transmitted from the pulley-side journal could create a localized stress maximum at specific angles along the keyway edge. The sharp corners of a traditional keyway act as potent stress raisers. The theoretical stress concentration factor \( K_t \) for such a notch can be approximated for preliminary assessment. For a rectangular notch under torsion, the factor depends on the geometry. A simplified expression relating to notch sensitivity in ductile materials like nodular cast iron is:
$$ K_t \approx 1 + 2 \sqrt{\frac{a}{\rho}} $$
where \( a \) is a characteristic dimension (like notch depth) and \( \rho \) is the root radius. A small \( \rho \) leads to a high \( K_t \), drastically reducing the fatigue strength of the component. The effective alternating stress \( \sigma_a \) experienced is then:
$$ \sigma_a = K_t \cdot \sigma_{a,\text{nom}} $$
where \( \sigma_{a,\text{nom}} \) is the nominal alternating stress. For the high-strength nodular cast iron used, fatigue failure can initiate at these stress concentrations even if the nominal stresses are within design limits.
We hypothesized that changing the force transmission element from a flat key to a cylindrical pin could improve the situation. The cylindrical pin would have the same diameter as the width of the original flat key. This redesign transforms the keyway from a long, sharp-cornered slot into a circular hole, which inherently has a better stress flow and a higher fatigue strength factor. The volume of material removed is also less, potentially increasing the net section strength of the journal. To quantitatively evaluate this, we conducted a comparative Finite Element Analysis (FEA).
Our FEA model simulated the front end of the crankshaft, applying boundary conditions replicating the assembly: an interference fit for the gear on the journal and a connection to the pulley on the outer diameter. We applied an equivalent bending load to the pulley journal to simulate operational stresses. The model for the nodular cast iron material used elastic-plastic properties derived from our tensile tests, with a Young’s modulus \( E \) of approximately 169 GPa and a Poisson’s ratio \( \nu \) of 0.29. The key comparison was the von Mises stress distribution around the keyway/pin hole under identical loading.
The results were revealing. For the flat key keyway structure, under a simulated downward bending force of 10,000 N applied to the pulley journal end, the maximum stress concentration occurred at the sharp edge of the keyway, reaching a value of approximately 75 MPa. In contrast, for the cylindrical pin structure, the maximum stress at the edge of the circular hole was significantly lower, around 60 MPa. This represents a 20% reduction in peak stress. The stress distribution was also more uniform around the circumference of the circular hole. The von Mises stress criterion, used to predict yielding in ductile materials, is given by:
$$ \sigma_{vM} = \sqrt{ \frac{(\sigma_1 – \sigma_2)^2 + (\sigma_2 – \sigma_3)^2 + (\sigma_3 – \sigma_1)^2 }{2} } $$
where \( \sigma_1, \sigma_2, \sigma_3 \) are the principal stresses. The FEA contours clearly showed lower \( \sigma_{vM} \) values for the cylindrical design in the critical region. This analysis confirmed that the modified design would alleviate the stress concentration responsible for initiating cracks in the nodular cast iron.
The improvement in fatigue life can be estimated using the stress-life (S-N) approach modified for stress concentrations. The modified endurance limit \( S_e’ \) for the notched condition is:
$$ S_e’ = \frac{S_e}{K_f} $$
where \( S_e \) is the endurance limit of the polished, unnotched nodular cast iron specimen, and \( K_f \) is the fatigue notch factor, which is generally slightly lower than the theoretical \( K_t \) due to the material’s notch sensitivity. For nodular cast iron, the notch sensitivity factor \( q \) can be defined as:
$$ q = \frac{K_f – 1}{K_t – 1} $$
By reducing the geometric stress concentration, the effective \( K_f \) for the cylindrical design is lower, leading to a higher \( S_e’ \) and, consequently, a longer predicted fatigue life under cyclic torsion and bending. The relationship between applied stress amplitude \( \sigma_a \) and cycles to failure \( N_f \) often follows the Basquin’s law in the high-cycle fatigue regime:
$$ \sigma_a = \sigma_f’ (2N_f)^b $$
where \( \sigma_f’ \) is the fatigue strength coefficient and \( b \) is the fatigue strength exponent. A lower \( \sigma_a \) due to reduced stress concentration directly increases \( N_f \).
Following the FEA validation, we implemented the design change. The flat key and its keyway were replaced with a cylindrical pin and a corresponding drilled hole. The manufacturing process for the nodular cast iron crankshaft was updated accordingly. Subsequent engine dynamometer tests were conducted on the modified crankshafts. The results were successful; the previously prevalent torsional fracture at the front small-end keyway was eliminated, and all test units passed the rigorous bench validation procedure.
In summary, our investigation into the failure of nodular cast iron crankshafts followed a systematic engineering approach. The initial comprehensive material characterization confirmed that the QT700-2 nodular cast iron met all specified requirements, eliminating material quality as the root cause. While initial machining imperfections were corrected, the failures persisted, redirecting our focus to the design geometry. Theoretical stress analysis and advanced FEA simulation quantitatively demonstrated that the traditional flat key keyway generated a significant stress concentration factor, making it a likely initiation site for fatigue cracks in the otherwise robust nodular cast iron. The redesign to a cylindrical pin connection provided a more favorable stress distribution, effectively lowering the peak stresses. This case study underscores the critical importance of integrating detailed stress analysis during the design phase of dynamically loaded components, even when using high-performance materials like nodular cast iron. The successful resolution highlights how a combination of empirical testing, metallurgical examination, and computational modeling can lead to effective engineering solutions for enhancing the durability and reliability of nodular cast iron parts in demanding applications.
The properties of nodular cast iron, such as its good castability, machinability, and excellent strength-to-weight ratio, make it a preferred choice for crankshafts. However, its fatigue performance is highly sensitive to stress raisers. The endurance limit \( S_e \) for nodular cast iron can be related to its tensile strength \( S_u \). A common empirical relation for ferritic-pearlitic grades is \( S_e \approx 0.4 \times S_u \) for rotating bending on polished specimens. For our QT700-2 with \( S_u \) around 800 MPa, this gives an approximate \( S_e \) of 320 MPa. The presence of a keyway drastically reduces this. The fatigue strength reduction factor \( K_f \) for the keyway can be estimated using Peterson’s equation:
$$ K_f = 1 + \frac{K_t – 1}{1 + \frac{a}{\rho}} $$
where \( a \) is a material constant related to the grain size and strength. For high-strength nodular cast iron, \( a \) is relatively small, meaning the material is more notch-sensitive, and \( K_f \) approaches \( K_t \). This explains the susceptibility to failure. The modified cylindrical hole, with a larger effective root radius, results in a lower \( K_t \) and hence a lower \( K_f \), preserving more of the inherent fatigue strength of the nodular cast iron.
Furthermore, the role of residual stresses from processes like fillet rolling should be considered. While our crankshafts undergo fillet rolling to induce beneficial compressive stresses in the main and pin fillets, the keyway area is typically not treated. The superposition of applied tensile stresses and the lack of compressive residuals at the keyway edge further exacerbate the condition. Future optimizations could explore localized surface treatments around the pin hole to introduce compressive residual stresses, potentially leveraging the good response of nodular cast iron to such treatments.
In conclusion, the transition from a flat key to a cylindrical pin connection in nodular cast iron crankshafts proved to be a decisive improvement. It addressed the fundamental issue of geometric stress concentration without compromising functionality. This solution enhances the structural integrity of nodular cast iron components subjected to complex multiaxial loading, ensuring reliable performance in advanced diesel engine applications. The methodology we applied—from failure analysis and material verification to computational simulation and design iteration—serves as a robust framework for solving similar engineering challenges involving high-integrity nodular cast iron castings.