The great grinding divide

Bridging the gap between academic and shop floor grinding knowledge.

Grinding is a “strategic process” that occurs close to the end of the production chain after much labor has been put into a product. As such, it can determine the success or failure of a product.

Because of its seemingly random nature and negative rake angles, grinding is often viewed as a difficult process. But significant technical advances have been made in recent decades, shifting the perception of grinding from a random to an understandable chip-formation process in the same vein as turning, drilling and milling. Unfortunately, though, much of the fundamental knowledge gained has not found its way onto the shop floor.

Courtesy of United Grinding Technologies

For example, in 1951, grinding researcher L.P. Tarasov clarified the vague definition of “grinding burn”—oxidation burn, thermal softening, residual-tensile stresses and rehardening burn—in hardened steel by describing the metallurgical changes that occur with each and the approximate temperatures when they occur. Sixty years later, however, confusion still abounds about what constitutes grinding burn, and engineers often still rely on visual oxidation to judge whether a workpiece has suffered thermal damage. Considering that oxidation burn begins at around 250° C, whereas genuine thermal damage typically occurs at 600° to 1,000° C and may be present in the absence of visual oxidation burn, using this visual method can prove dangerous.

Figure 1 (below) shows oxidation burn after thread grinding (on the nonground surface in the flute-grinding region). Temperatures in this oxidized area were much lower than in the hot-spot region on the thread-ground surface—the clean surface. However, in the thread-ground region, the oxidation burn was ground away, whereas it was not on the nonground surface.

There is no way to determine if genuine thermal damage is present without using a more involved testing procedure, such as polishing, etching, X-ray diffraction or acid cooking. However, many operators and engineers believe that if the tool is clean, it isn’t burned, and if it’s brown and blue, it’s burned. That belief is false and risky.

The following are other key examples of fundamental grinding concepts that are often not well known by production engineers.

• Coolant velocity should match wheel velocity, with high flow rates not being necessary.

• Rotary dressing in the antidirectional mode dulls the wheel, whereas dressing in the unidirectional mode with a speed ratio (v_dresser/v_wheel) of 0.8 produces a sharp wheel.

• In single-point dressing, the dressing lead and overlap ratio are far more important in determining wheel sharpness than the dressing traverse speed.

• Diamond wheels go out of true if they are trued on an adaptor, taken off and remounted on the machine spindle.

Because these basic concepts haven’t made it to the shop floor, grinders are learning them the hard way—through trial and error.

Applying Knowledge

I have visited grinding facilities in 29 countries and have seen low-tech grinding of drill bits and high-tech grinding of artificial-knee implants and turbine blades for jet engines. Regardless of the country or industry, many companies do not understand the basic concepts of grinding and, more importantly, do not have access to these concepts presented in an easily accessible, practical way.

I call this “The Great Grinding Divide” because there is a large gap between the knowledge held in academia and the knowledge held on the shop floor. One reason for this divide is that basic concepts have not been translated into a simple, useful format that can be quickly utilized by the grinding machine operator.

For example, as opposed to turning, which can be readily modeled in two dimensions, the 3-D nature of grinding makes calculating chip thickness difficult. Machine operators accustomed to the speeds-and-feeds diagrams used for turning are frustrated that such a relationship does not exist for grinding.

The most common equation for maximum chip thickness in grinding, h_m, is some variant of:

where C is the cutting-point density, r is the shape factor, v_w is the workpiece velocity, v_s is the wheel velocity, a_e is the DOC and d_e is the equivalent diameter.

However, measurement of C and r is rather subjective. More importantly, the equation is intimidating even to those with a higher education who work solely in machining.

Fortunately, the equation can be simplified to a speeds-and-feeds equation that uses only the machining parameters that can be varied—DOC, feed rate, wheel speed and wheel diameter—and called “aggressiveness.”

But this equation still uses variables that must be defined, along with units, and has typical aggressiveness values that are small, such as 1.8×10^-5.

Therefore, the equation can be further simplified by multiplying the resulting value by a constant of 1 million to give reasonable values.

The issue of variables, nonetheless, still exists, along with the issue of units and unit conversions. Therefore, we can rewrite this equation in friendlier terms as:

With this equation, the machine operator can plug in the values from the CNC and generate an understandable number.

Typical aggressiveness values are from 3 to 60, with lower values for finishing and higher values for roughing. Moreover, each combination of wheel, workpiece and coolant will have an optimal aggressiveness value that will place it in the “sweet spot” of the operation, the place where the maximum chip thickness is large enough to form a chip and avoid excessive rubbing and high specific energies, but not so large as to cause excessive wheel wear.

This is an example of taking a complicated concept and modifying it to provide a simple yet useful parameter. Machine operators can identify with the concept of aggressiveness, and there is no need to labor over complex variables and unknown units; it’s all given in a format that can be entered into a calculator.

This begs the question: Why haven’t similar concepts been translated into similar, easy-to-use techniques? And if they have been, is this information accessible to those on the shop floor?

No Access to Basics

An even simpler formula that’s used in calculations in just about every grinding process is the specific material-removal rate, Q’, known as “Q-prime.”

Q’ = a_e × v_w

where a_e is the DOC measured in mm and v_w is the feed rate measured in mm/sec. The specific mrr is the total mrr per unit width of the grinding wheel.

Unfortunately, many companies are not aware of this calculation. For example, a company I visited in Europe was trying unsuccessfully to cylindrical grind hardened steel with a CBN wheel using a Q-prime value of 82. An ambitious Q-prime for that application is around 15. This company was trying to make it work with 82 and madly adjusting the dressing and cooling parameters and wheel speed to no avail.

Why was the engineer on the project trying to make a process work with parameters that were outside the practical realm of possibilities for this wheel? Because he was missing a piece of the puzzle. The engineer had a degree in mechanical engineering and was capable of high-level math, but he did not have access to the concept of specific mrr or reasonable values for cylindrical grinding. He was investing his energies in cooling and dressing when he should have been focused on grinding parameters. Once the concept was explained to him, he quickly adjusted the parameters to a more realistic Q-prime value.

One reason that companies lack people with grinding knowledge is these people have retired or were laid off during the cutbacks in the 1980s and in the era of “lean manufacturing,” which occurred mostly in the 1990s. Many of these people were not replaced, and their knowledge disappeared with them.

Figure 1. Oxidation burn from thread grinding.

In 2002, I visited a multinational company working in grinding and was impressed by the workers’ high level of expertise and the company’s advanced research program, which included numerous test machines and measuring devices.

Six years later, the company asked me to give my basic 3-day grinding course and provide general technical advice about how to set up a test program. Not a single person from the 2002 group was still at the company, which had completely lost its technical expertise.