Philosophical Musings about Grinding Part 1/3

PART 1/3

This essay is divided into three separate LinkedIn publications. As a whole, it looks at different approaches to establishing grinding processes. While I proudly work for Reishauer, a gear grinding machine tool builder, I wish to point out that the following ideas are my private musings. However, I do not doubt that my coworkers, who will without any doubt read this, will bring me down to earth and put me in my place if I have gone too far off on a tangent.

The ideas expressed here only touch on gear grinding and focus more on my experience across various grinding processes such as cylindrical grinding, creep-feed, and centerless grinding, among others. I believe that all grinding operations share a lot in common and understanding one process in depth is an open gateway to understanding other methods in their commonalities and differences.

Furthermore, the intention is not to provide a scientific treatise. Still, I hope my ideas are, practically speaking, anchored on firm engineering grounds and prove to be useful in a workshop environment. As I have stated at the outset in the title, these are philosophical musings. Hence, I may at times be excused for being a bit opinionated. For this reason, my readers may forgive me when, at times, I stray off the topic of pure grinding, particularly at the beginning when I muse on attitudes and reasoning.?

The three parts of the essay are established on a timeline from past to present to the future. The first two parts deal with the past to present on how we have found and do establish a process that can be deemed efficient. What methods were and are at our disposal to set up an efficient grinding process? More interestingly, in my view, is what the future will look like for establishing grinding processes. In Part 3, I look at how we could harness the power of artificial intelligence (AI) applied to grinding processes. As shown in Illustration 1, I believe that AI will revolutionize grinding processes and greatly increase our knowledge of grinding.?

Illustration 1: Grinding knowledge and artificial intelligence (AI)

This essay is subdivided as follows:

1.?????What attitudes should an engineer have?

2.?????A short discourse on the use of reasoning

3.?????A practical approach to setting up the grinding parameters (Reduce to the max!)

4.?????Using computer modeling and force measurement in grinding processes

5.?????The future of grinding using big data and artificial intelligence (AI)

1.?????Attitudes of Engineers

At the age of 16, I began a toolmaker’s apprenticeship in Switzerland that was the foundation of the next 50-plus years I have spent in engineering. You can infer that I’m an old man in years. However, in spirit, I remain a boy. During my apprenticeship and after, I met people that were true masters in their field. They took me under their wings as I was curious and wanted to learn, and they often gave me a hard time when I did not live up to their expectations. I’m still grateful for this, as it pushed me to improve myself. This has continued; I’m still meeting people who know a lot more than me, and I continually exchange ideas with my peers. I’m still curious, willing to learn and change, and ready to admire others for their knowledge.

Interestingly, these attitudes make one unthreatening to others, and people open up and are willing to share their knowledge. Curiosity is an essential element of any engineer, coupled with a healthy dissatisfaction and a degree of unreasonableness concerning the present state of any machine, process, or gadget. To quote George Bernard Shaw, who summed it up more succinctly than I could:

“The reasonable man adapts himself to the world; the unreasonable one persists in adapting the world to himself. Therefore, all progress depends on the unreasonable man.”

Going back to curiosity, this is best explained by a joke, the origin of which escapes me:

When I was a child, my father asked me what I would like to be when I was grown up. I replied: “When I’m grown up, I want to be an engineer.” To which my father calmly added: “You cannot be both!”

2. Rational Thinking

One should assume that in the field of engineering, rational thinking dominates. But is this true? Are engineers completely rational people? My experience has been that rational thinking does not always prevail. The English philosopher and mathematician, Bertrand Russel, used a harsher judgment than I would:

Man* is a rational animal. So at least we have been told. I have searched diligently for evidence in favor of this statement throughout a long life. So far, I have not come across it.”

(*Contemporary wording would suggest the terminology of “human being” rather than “man.”)

There are reasons not to side with Bertram Russel and to take heart from a study of preschoolers?1)?who showed clear and innate abilities for logical deduction and rational thinking, which may counter Russel’s argument. The experiment went like this: A man dressed up in a monkey suit appeared three times on a stage, holding two different colored flowers.

Illustration 2: Study of preschoolers

In total, the monkey had three flowers: red, yellow, and blue, giving three combinations of two different colored flowers. Only one of the three flowers caused an allergic reaction and made the monkey sneeze.?

Illustration 3: Reactions to three pairs of different colored flowers

79% of 4-year-old preschoolers determined the cause of sneezing: The blue flower.

Conclusion of this study: The ability to tell cause from effect does not require specific instruction and, therefore, is innate (we are borne with it).

Why I’m I telling this story? By their training and inclination, engineers are a group of people who can decide rationally. Well, unfortunately, this is not always the case. When selling grinding wheels, I came across a lot of non-rational bias and unwarranted resistance to change. For example, many years ago, it was my job to replace a white aluminum oxide grinding wheel specification with a new version of the same abrasive. The version to be replaced featured old bond technology, whereas the new version featured recrystallized glass bonds with superior grinding properties, even though the abrasive grain remained the same. The problem was that the old version used iron oxide to color the wheel brown, and the new version used chromium oxide to color the wheel green. While numerous tests had proven the “green” wheel superior to the “brown” wheel, the application engineers at a Swiss grinding machine tool builder refused to change over as they “felt” the green wheel could not match the performance of the brown wheel they were used to. As chance would have it, an important job needed to be ground at some stage, and only a green wheel was instantly available. Lo and behold, the green wheel performed well beyond their expectations. The same people who resisted change over to the new wheel technology suddenly became converted and turned into our best ambassadors. I had a similar experience with introducing ceramic wheels (Cubitron 321). Seemingly rational people refused to test the revolutionary grinding technology, usually on flimsy grounds. Initially, it took much effort to bring them around to see the benefits. In short, engineers do not always act as rationally as they believe they do, and I could produce an exhausting list to prove the point. Hence, introducing new approaches to grinding often requires a person who’s more psychologist than an engineer.?

For those readers who would like to pursue the idea of rationality – and the resistance to it – I recommend the book “Rationality” by Steven Pinker?2), a cognitive scientist and professor of Psychology at Harvard University. A small intro from the cover of the book:

“In the twenty-first century, humanity is reaching new heights of scientific understanding – and at the same time appears to be losing its mind. How can a species that discovered vaccines for Covid-19 in less than a year produce so much fake news, quack cures, and conspiracy theories?”

In “Rationality,” Pinker rejects the cynical cliché that humans are simply an irrational species – cavemen out of time fatally cursed with biases, fallacies, and illusions. Instead, he explains, we think in ways that suit the low-tech contexts in which we spend most of our lives but fail to take advantage of the powerful tools of reasoning we have built over millennia: logic, critical thinking, probability, causal inference, and decision-making under uncertainty.

3. A practical approach to grinding

Following the small digression about rationality, how can we apply this human quality to grinding? One of the problems of establishing grinding processes is the huge number of parameters that define the grinding process. According to VDI (Association of German Engineers), a grinding process has more than 140 parameters, most of which we cannot know unless we have a laboratory machine for which all features such as stiffness of spindles, bearing machine base, etc. have been established. Even then, as we may have a great range of workpieces, how about the stiffness of those and that of the varying clamping fixtures? What about the huge range of grinding wheel specifications available? How about the cutting edge density of the grinding wheel, the dynamic displacement, and other parameters that torture our brain, as shown in Illustration 4?

Illustration 4: Grinding parameters difficult to know

The above picture illustrates the mental pain caused by so many unknown factors. Also, I often wondered, do we engineers enjoy complexity when we should be looking for simplicity? The following formula of the Work Removal Parameter (WRP) seems a case in point of unnecessary complexity and perfectly suited to scare off any eager practitioner of grinding. How could one arrive at an exponent of De?43/304? This formula suggests control over the grinding process, which I judge as illusory.

Illustration 5: Complex formula of a grinding parameter

The approach I have taken was taught to me some 30 years ago by a Swiss grinding pioneer, H. W. Ott?3). I would call the approach “Reduce to the Max” and “Near enough is good enough,” which is to say choose the parameters you know and which you can calculate by simple formulas and for which guidelines exist. The guideline can be taken from the images. One can easily make an Excel worksheet to calculate grinding parameters with the guidelines below. The result is never perfect, nor would I ever look for perfection as it is illusory. Hence, near enough is good enough, and the rest is iteration. Six of the most important parameters are:

Q’w???????Specific material removal rate in mm3/mm/sec

qs?????????Speed ratio between grinding wheel and workpiece

hm????????theoretical chip thickness in?mm

CL????????Classification of the coolant and its lubrication factors (acc. to Ott)

MA???????Material classes and grindability (acc. to Ott)

Fa?????????Aggressiveness factor of a grinding wheel (acc. to Dr. Jeffrey Badger)

The first four parameters can establish a grinding process that is “good enough.” The last two factors, CL and MA, we will encounter in more depth in Part 2 of this essay. Finally, it should be mentioned that several factors not listed here, such as the surface speed vc, the workpiece speed vw, the depth of cut ae, etc., are used in the calculation of the above six factors.?

Illustration 6: Specific material removal rate, aka Q-Prime

Illustration 7: Specific material removal rate, aka Q-Prime

Illustration 8: Speed ratio between grinding wheel and workpiece

Illustration 9: Theoretical average chip thickness

Illustration 10: The aggressiveness factor Fa

The specific material removal rate Q’w?(also known as Q-prime),?explained above in Illustrations 6 & 7,?is used to evaluate the performance of a grinding process. Q-prime indicates how many mm3?per one (1) mm wheel width removes per second (mm3/mm/sec). Two parameters are needed to calculate Q-prime: The depth of cut ae?and the feed-rate vw.?

If the intention is to increase the material removal rate, the question arises whether it would be better to increase the feedrate vw?or the depth of cut ae. This question illustrates the limitation of using Q-prime (Q’w) as a benchmark as the depth of cut ae?and the feedrate vw?are equally weighted. A benchmarking concept that goes beyond the mentioned limitation was developed by Dr. Jeffrey Badger (www.TheGrindingDoc.com ).?Based on these parameters, an “aggressiveness factor” is calculated, which allows establishing if a grinding process runs in its optimal range if allocated to specific grinding operations and specific grinding wheels. In this description, this aggressiveness factor is called “Fa.”

Once an optimal Fa?has been established, it can be utilized for process optimization, irrespective of whether grinding the wheel diameters and the wheel speeds are different from one process to another. Fundamentally, the aggressiveness factor Fa?refers to the chip volume and the chip thickness hm, and their influence on the process behavior of a given grinding wheel.?Readers who want to know more about the aggressiveness factor Fa can consult my LinkedIn article on this very topic. To give but one example, I have used the aggressiveness of Fa?at a factory that produces high volumes of twist drills that had their flutes creep-feed ground in one pass. They used three different specifications of grinding wheels. Once we had established each grinding wheel’s aggressiveness factor Fa, we could use the same spec at different diameters, for example, and tweak the surface speed and feed rates in such a way, that we could grind at the previously established aggressiveness factor Fa. At this junction, it’s worth mentioning that the aggressiveness works best with surface and creep-feed grinding operations, including gear grinding. For cylindrical grinding, due to the small arc of contact, the Fa?factor is, unfortunately, of lesser usefulness. The following is a guideline for Fa?as applied to different processes:

• Profile gear with single rib wheels??????????????????????????25 to 35
• Generating gear grinding???????????????????????????????????40 to 50
• Creep-feed grinding??????????????????????????????????????????15 to 20
• CD-creep-feed grinding?????????????????????????????????????20 to 25
• Reciprocating surface grinding?????????????????????????????????60 to 80

CL and MA

The next two parameters – CL and MA – are not needed for a simple process calculation, as discussed in Part 1. However, as we dive deeper into detail in Part 2/3, the two factors become important when looking at computer-assisted programs. The CL sc?factor is derived by dividing the tangential force Ft?by the normal force Fn. The normal force should always be much larger than the tangential force as high tangential forces lead to grinding burns. High normal forces imply an increased pressure on the individual abrasive grain, which leads to self-sharpening and thus to an aggressive grinding process and a lower risk of burning. Ideally, the CL factor should be < 0.5.

Illustration 11: Coolant lubricating properties of different media

The range of the coolant lubricating factor sc?is as follows:

Dry grinding????????????????????????????????????????????????????? ?CL = 0

Water??????????????????????????????????????????????????????????????? CL = 1

Soluble with 30% mineral oil in concentrate???????CL = 3

Soluble with 40% mineral oil in concentrate?????? ?CL = 3.5

Neat oil with EP additives?????????????????????????????????? CL = 5

The above range can be fine-tuned in the grinding application programs for additional coolants. In the second part of this article, the fine-tuning will be explained.

Material classes MA

Illustration 12: Types of steel and grindability

The steel material classes (MA) feature the following range within the application grinding programs. Each steel class has a specific attribution value that influences the specific material removal rate (Q’w) that the program will suggest.

Very easy to grind (mild steel)?????????????????????????? ?1

Easy to grind constructional steel????????????????????? ??2

Easy to medium to grind steel??????????????????????????? ?3

Medium to grind low alloy steel????????????????????????? 4

Medium to difficult to grind alloy steel??????????????? 5

Difficult to grind alloy steel???????????????????????????????? 6

Very difficult to grind high-alloy steel (Cr, W)??????7

This concludes Part 1/3. In the next part, Part 2/3, we look at computer-assisted grinding programs based on the simple premises as shown in Part 1 but containing some added features such as calculating grinding forces, coolant requirements, etc.

Walter Graf, The Philosopher’s Grindstone, Copyright???April 2022

References

1.?????Alison Gopnik, Laura Schulz; Causal Learning: Psychology, Philosophy, and Computing; Publisher: Oxford University Press, March 2007

2.?????Steven Pinker, Rationality: What It Is, Why It Seems Scarce, Why It Matters, Publisher: Allen Lane; 1st Edition 28 Sept. 2021.

3.?????H.W. Ott, Schleiftechnik, 1997 to 2008, self-published.