In our investigation into the precision grinding of hardened gear tooth surfaces, we identified grinding burns and grinding cracks as critical defects that compromise gear performance and lifespan. Grinding burns occur due to localized high temperatures (900°C to 1500°C) during the process, altering surface microstructure and causing oxidation. This often leads to grinding cracks when residual stresses exceed material limits, resulting in micro-fractures that accelerate fatigue failure, pitting, and system contamination in gear transmissions. Our research focused on optimizing grinding parameters while preventing these defects, particularly by monitoring overload currents in form grinding machines to resolve the conflict between efficiency and crack prevention.
Grinding cracks stem from multiple factors, including inadequate heat treatment in prior processes, such as excessive surface hardness or carbide networks in carburized gears. Preparation aspects like improper grinding wheel selection (e.g., material, hardness, grit size, porosity) or insufficient cooling fluid exacerbate the risk. To address irregular thermal distortions from hardening, we implemented real-time current monitoring to detect instantaneous overloads during initial grinding phases. This approach allows dynamic adjustment of feed parameters without relying on operator judgment, reducing unproductive “air passes” and cutting machine time. The core relationship involves grinding force (F) and motor current (I), expressed as:
$$ F = k \cdot I $$
where k is a machine-specific constant. Overload currents indicate excessive cutting forces, which correlate with temperature spikes that cause grinding burns and grinding cracks. We validated this through experiments on carburized gears with asymmetric tooth profiles and root grinding requirements. Key variables included radial depth of cut, stroke speed, and grinding stages (roughing, semi-finishing, finishing).
Our experimental setup involved a high-precision form grinding machine, with grinding cracks detection via acid etching and magnetic particle inspection. We tested two wheel brands—Norton and COMET 3SG60—under varying current limits. The grinding process was divided into stages: roughing used double-sided grinding for bulk removal, while semi-finishing and finishing employed single-sided grinding for accuracy. Critical parameters are summarized below:
| Grinding Stage | Wheel Brand | Grinding Type | Overload Current Limit (%) | Key Parameter Adjustments |
|---|---|---|---|---|
| Roughing (Passes 1-3) | Norton 3SG60 | Double-sided | 10 | Standard depth and speed |
| Semi-finishing (Passes 4-5) | Norton 3SG60 | Single-sided | 13 | No radial feed |
| Finishing (Pass 6) | Norton 3SG60 | Single-sided | 8 | Focus on surface finish |
| Roughing (Passes 1-2) | COMET 3SG60 | Double-sided | 18 | 100% depth increase, 40% speed increase |
| Semi-finishing (Pass 3) | COMET 3SG60 | Single-sided | 15 | 20% parameter increase |
| Semi-finishing (Pass 4) | COMET 3SG60 | Single-sided | 18 | Maintained accuracy |
| Finishing (Passes 5-6) | COMET 3SG60 | Single-sided | 15 | Unchanged for roughness |
With Norton wheels, setting current limits at 10% for roughing prevented grinding cracks while maintaining 100% product quality. However, COMET wheels allowed aggressive optimization: roughing at 18% current limit with doubled depth and 40% higher speed reduced grinding time by half. This also lowered wheel consumption from 0.7 mm per gear to 0.5 mm and extended diamond dresser life by 33%. The heat generation during grinding can be modeled to predict grinding burns risk:
$$ T_{\text{max}} = \frac{q \cdot v_s}{h \cdot a} $$
where Tmax is peak temperature, q is heat flux, vs is wheel speed, h is heat transfer coefficient, and a is contact area. By keeping I below thresholds, we controlled q to avoid critical Tmax values that induce grinding cracks.
Post-grinding inspections confirmed zero grinding cracks across all samples, validating the current-monitoring approach. Efficiency gains were quantified through reduced cycle times and resource usage, as shown in the comparative table below:
| Performance Metric | Original Process | Optimized Process (COMET) | Improvement (%) |
|---|---|---|---|
| Pure Grinding Time | 100% (baseline) | 50% | 50 |
| Wheel Consumption | 0.7 mm per gear | 0.5 mm per gear | 29 |
| Diamond Dresser Life | 100% (baseline) | 133% | 33 |
| Defect Rate (grinding cracks) | 10% | 0% | 100 |
This method not only mitigates grinding cracks but also enhances economic viability by lowering costs. For instance, the stress intensity factor for grinding cracks propagation is:
$$ K_I = Y \sigma \sqrt{\pi a} $$
where KI is stress intensity, Y is geometry factor, σ is applied stress, and a is crack length. By minimizing σ through controlled currents, we suppress crack initiation. Future work should integrate automated feedback systems to further reduce grinding cracks incidence, as post-grinding checks remain essential for critical gears.