Bottlenecks in Gear Technology

Based on personal experience, I divide the bottlenecks encountered in the gear industry into four stages.

Understanding and mastering gear parameters.
Having a complete gear inspection system.
Applying general-purpose modification technology.
Applying specialized modification technology.

The first stage is understanding all parameters involved in gear design and manufacturing, being able to correctly complete gear design, and manufacturing gears according to those parameters.

This bottleneck is relatively easy to overcome. Currently, it mainly relies on gear design software. The parameter tables exported by professional software basically cover all gear parameters and have already considered potential issues. As long as you use professional software correctly, this is not a bottleneck. And the quality improvement before and after this breakthrough is very obvious.

The breakthrough path is relatively simple — read gear books and articles, purchase professional software, and learn to use it correctly.

The second stage is having a complete gear inspection system. Many domestic companies, to cut costs, may use software for gear design and then outsource manufacturing. Once they receive the gears, they use them directly. When these companies encounter poor meshing performance, they typically assume the design is wrong, then keep tweaking parameters and re-manufacturing. I've seen many such cases.

Case 1: A company had gear noise and wanted to redesign. I verified the parameters — the design was actually fine. I suggested it might be a manufacturing issue and recommended measuring the profile and lead first. The customer replied that their supplier was a listed gear manufacturer, so manufacturing problems were unlikely. Still, they obtained an inspection report. The report revealed that the manufacturer had applied modification — that was fine — but the manufacturing precision was appalling. Clearly, no mid-process inspection had been conducted; the profile deviations were outrageously out of tolerance, yet they shipped directly to the customer. When my customer confronted the supplier, they acknowledged the poor quality but refused to fix it. Likely a large company that didn't care about my customer's small volume. The issue was finally resolved by switching suppliers. So would redesigning help in this scenario? Actually, in this case, perhaps yes — because the supplier had the technical capability, and early-stage improvement would certainly be possible. The real problem was they didn't value this customer. Not their core client, so mass production was careless, and problems would recur. If measurements had been taken, the issue would have been detected early, rather than buying a batch of defective products and then hiring someone to analyze them.

Case 2: A long-term customer of ours used our design parameters to manufacture gears with obvious noise. They reported smooth hand-meshing but excessive noise after installation. When we received the gears, we could feel the binding during hand-meshing. Measurement revealed incorrect pressure angle on the internal gear. The gear cutter manufacturer admitted the error and provided new tooling. This customer also suspected parameter design issues and wanted to redesign, rather than suspecting the cutter manufacturer had made the pressure angle wrong. Had we not insisted on measuring the actual parts, the customer would have borne all the losses. With proper measurement, the trial-cutter process would have caught the problem, and they wouldn't have manufactured a whole batch before discovering the issue.

Such cases are extremely common. Most problems we solve are caused by manufacturing or assembly issues resulting in poor gear meshing quality. Yet these customers all believe their parameter design is at fault. Their common characteristic: they lack proper gear inspection capabilities. Whenever a problem arises, they redesign and re-manufacture — like restarting a computer when it has problems. Re-manufacturing might eliminate the original manufacturing defect, and they credit the parameter design. If it's still not fixed, then "the parameters aren't right yet — restart again."

If you've already mastered correct gear parameter design and want to advance further, you must build physical gear evaluation capabilities. Design cannot solve manufacturing problems. Assembly and manufacturing issues are the primary causes of gear noise and failure in the cases our company encounters.

The solution to this bottleneck is spending money — on measuring equipment and quality inspection processes. Breaking through this bottleneck enables you to consistently produce qualified products, eliminating the inconsistency of some products being good and others poor.

The third stage is applying general-purpose modification technology. Its characteristic: it basically fits all gears — once applied, gear performance improves. I consider modification as useful as the lead-in chamfer in shaft-hole assembly. Those who say modification is useless are like saying shaft-hole assembly doesn't need a lead-in chamfer. I simply don't believe that. How to modify is like how to design the lead-in chamfer shape. A poorly designed chamfer doesn't mean chamfers are ineffective.

Why does modification sometimes seem ineffective? Because you haven't understood its purpose. Representative general techniques are short profile modification and equal-radius lead crowning. Modification amounts follow recommended standard values adjusted by practical experience. No complex calculations, but excellent results. This modification theory's core is reducing the adverse effects of manufacturing and assembly errors. It should primarily be applied where manufacturing/assembly errors are much larger than deformation-induced errors. The fundamental purpose is not to raise the upper performance limit, but to ensure mass production consistency. Many companies face cost constraints where gear and housing assembly errors can only be so large. In mass production, a small percentage inevitably performs poorly. After applying general modification, the proportion of underperforming products drops dramatically; with stable processes, nearly 100% pass rate is achievable. At this stage, every product in mass production has relatively stable and reliable quality.

Stage four: specialized modification technology. There is no unified algorithm — each gear pair requires individually designed modification. It's generally based on load deformation, temperature changes, and other factors. The purpose is to raise the upper performance limit. General modification appears somewhat crude in comparison, since its curves are regular lines or arcs. Specialized modification aims to produce a curve that minimizes mesh-in and mesh-out impact. It must adapt to the minute changes in the meshing tooth surface. Because specialized modification is optimized for these minute surface changes, it requires extremely high manufacturing precision to demonstrate its power. With insufficient precision, the optimization only benefits some mass-produced products while others may see no benefit or even negative effects. This is the fundamental reason why analysis-and-simulation-based modification often shows little improvement in mass production — your company's production precision is average, so only some products reach a higher ceiling while others don't match the modification curve. So, managers and engineers: when someone talks about modification and always brings up deformation theory, ask yourself whether your products achieve 100% mass production pass rate. If yes, you have the potential to advance to this stage. Otherwise, wait — your products still need to address manufacturing and assembly errors as the primary concern. Deformation error is not your problem yet.

The figure below shows a lead modification chart for a 21,000 kW gear with 148 m/s pitch line velocity. Values are in micrometers.

Companies that master stage four technology are essentially at the top of the industry. This stage isn't about just having professional analysis software — it also requires the ability to calibrate software models to match reality.

I wrote this article because I frequently encounter manufacturers who haven't even reached stage one trying to use stage four technology. You have neither the manufacturing capability, the inspection capability, nor the application conditions — the technology you seek is worthless for your products.

I hope business leaders can accurately assess their current capabilities and advance step by step. This is the way to improve product quality at minimum cost and gain competitive advantage. Otherwise, the effort is wasted.

Presented by ETAGEAR

9/22/2023 8:00:00 PM