What is Fatigue Failure?

Let’s say you are leaving the house and you see your new credit card on the counter. You replace the expired card in your purse or wallet, but there are no scissors to cut up the old card. So, you start bending it back and fort until it breaks in half; then pitch it. You have just induced a fatigue failure.

So too with steel. A piece of steel that undergoes repeated motion (twisting, bowing, vibration, flex) will at some point fail. Conditions may encourage that failure to be earlier than expected. A nick or gouge at the surface, or an inclusion or defect (foreign element) within the steel may be the likely culprit.

In my experience, most industrial steel shaft failures are caused by fatigue. The failure may begin at the surface of the shaft (surface initiated), or, it may begin from inside (internal). Most common, will be surface initiated. Surface nicks are called “STRESS RISERS”. Think “A chip in the windshield of your car.” If not smoothed out, that nick will eventually become a crack that runs outward until the windshield fails. A Stress Riser is a break in the surface continuity of an item. Through repeated external forces, the surface-initiated nick becomes a fracture; progressing internally through the steel until a point of catastrophic failure. The shaft cracks in half, while the machine is still running. Not good!

If there are inclusions within the metal (microscopic tramp elements), they may contribute to an internally initiated fatigue failure.

What can we do to minimize fatigue type failures? First and foremost, make sure you are working with high quality materials. In the case of steel, make sure it has a high degree of cleanliness, free of internal defects. There are methods of steel production that insure your steel has excellent core integrity. Those are generally referred to as Clean-Steel-Production-Methods. Those methods, such as; Melting in an electric furnace, vacuum degassing, inclusion shape control, stirring, limiting tramp elements, etc. are available to people who require steel that has undergone refining processes. Certainly, you would want to employ those processes for materials that would be used in critical service.

Those processes address the internal portion of the steel. What about the surface? You can process the steel so as to minimize any roughness, nicks or gouges on the surface. If it is a shaft, you may want to insure you have a polished surface finish, even if the tolerance requirements of the application do not require a precision finish. Note that a highly polished surface not only resists surface-initiated fatigue failure, it provides a certain degree of corrosion resistance. Caution should be observed so that you do not get such a smooth surface that required lubricants will not adhere to the shaft. Most commercially available polished shafting will have a surface finish of about 15 micro. When you start getting into finishes much brighter than that, you may want to check into lubrication requirements.

If your finished part has contour changes (keyways, step-downs, grooves, etc), make sure the sharp corners have been radiused and if possible, even smooth out the contour.

FOOD FOR THOUGHT; “Most heavy industrial shaft failures are fatigue related. Toughness resists fatigue failures. Clean Steel Production Increases Toughness.”

-Howard Thomas, December 3rd 2018



Surface Hardening vs. Through Hardening

In the world of heavy industrial maintenance steel, whether you call it Case Hardening, or, Surface Hardening, or, Skin Hardening, it is all the same thing. This is a localized method of hardening employed to develop a wear resistant surface while maintaining a somewhat ductile (shock resistant) core. With production items, such as gear teeth, this may be very fine tuned, sophisticated, accurately measurable. In maintenance, “one-off” items, it can be somewhat erratic and capable of surprise. If you are contemplating increasing the surface hardness of a piece of steel, please recognize that increasing the hardness, especially localized hardness, is also increasing the brittleness which subsequently increases the chances of facture. Wear appropriate safety gear. Any surprises may not be very forgiving. 

IN GENERAL, the two types of hardening are self-explanatory. A through hardened piece of steel is pretty much the same relative hardness from surface to core. Most common prehardened steels, carbon or alloy, are often shipped at a hardness of approx. 30RC. As the cross sections get larger, the hardness will “drop-off to core”. That is, as you get closer to the center of the mass, the hardness may drop a few points. Those are still considered to be Through Hardened. Surface hardened levels, typically those used in hydraulic applications, and precision automation rail applications, will be supplied with a very thin hardened surface “skin”, at about 60RC, with a great drop off in hardness toward core.

The surface hardened material provides great resistance to sliding abrasive wear while resisting bending and torque. The through hardened alloy or carbon material provides a good balance of toughness (a combination of wear, impact, and gouging resistance). The through hardened material makes no pretense to be particularly ductile. In fact, as through hardness increases, the potential for general fracture also increases.

Caution should be exercised when attempting to surface harden small cross sections. Even though your intent and processing method may be aimed at surface hardening, small cross sections cool rapidly. The rapid cooling may actually result in a through hardened condition with potentially dangerous brittle hardness.

-Howard Thomas, Nov 8th 2018

Bearing Quality Vs. Bearing Fit

Be careful, the two are often confused between end-user and vendor. They are not interchangeable. BEARING QUALITY refers to manufacturing restrictions that are employed, when the mill makes the steel. It is commonly referred to as “Clean Steel Production”. The processing refines the steel, removing non-metallic inclusions and generally improving the quality and core integrity of the steel. This is not something that can be achieved in subsequent processing; it either is bearing quality, or it isn’t.

BEARING FIT (TOLERANCE), is achieved by subsequent processing of a steel bar or shaft. You may accomplish this at any point prior to use of the bar; do-able on-site locally, or as a specification for subsequent processing accomplished at the mill. This, as stated, refers to the tolerance alone; making no statement in regards to the integrity nor the cleanliness of the material. Bearing Tolerance is referred to as a “minus/minus” tolerance, as opposed to a plus or minus tolerance.

Typically, and depending on diameter, the tolerance would be something like minus .001″ to minus .0015″

A NOTE ON EXPRESSING BAR TOLERANCES, it is common to hear bar tolerances specified as; “Plus nothing minus .002”, or whatever the downside tolerance is. To avoid potential problems, it is better to state both plus and minus terms with a specific decimal position. Make sure that both parties know how far out that “plus” side is carried. Is it measured to the third decimal place, or the fourth? Many “Plus Nothing” bars actually may only be measured to the third (thousandths) place which allows the bar to actually be a plus tolerance. Such as plus .0005″. Note that Bearing Fit or Bearing Tolerance insures a minus tolerance.

You may further avoid potential headaches by specifying the actual diameter wanted in terms of the actual decimal, both over and under (plus and minus), such as .2500″/.2495″.

Whether it is Bearing Quality or Bearing Fit, keep in mind that there may be an additional cost to obtain that benefit.


-Howard Thomas, Oct 18th 2018

Centerless vs. On-Center Grinding

Bar Grinding Centerless Vs. On-Centers – Second Part of Four Part Set

As we mentioned in our last blog; in the maintenance industry, if someone refers to grinding a steel shaft, they are most likely talking about “Centerless Grinding”. There is another method, however, and that method is called “On-Center Grinding”. A misunderstanding on which method is actually required usually results in expensive errors, and general unhappiness for all parties. Of the two types, centerless is by far the most common. So much so, that if you mention grinding a shaft, the mill or service center will assume you are discussing centerless grinding.

Centerless grinding tends to follow the outside diameter of the bar; think apple peeler. When the skin is off, you still have a recognizable apple; naked, but still looks like an apple. Grind an egg-shaped hot rolled bar, and you will have a precision finished egg. In the hands of an experienced grinding operator, many troubling issues may be corrected. Taking it to an art form, the right operator can minimize irregularities and even affect straightness; to a point. The standard in industry is centerless. So, unless specified, tolerances being discussed are taken to be based on centerless.

On-Center grinding, on the other hand, indexes on the center of both ends of the bar. The grinding head then machines the O.D. of the bar to be concentric with the I.D. (chucked up centering holes). If your bar is egg shaped, now, your ground bar will be concentric. If the bar is bent, the finished ground bar will be straight, depending on how bent it was and how much stock removal you are able to take. The roundness (concentricity) and the straightness come from the “On-Center” grinding. On center grinding requires more stock allowance “to-clean up” than centerless grinding. Where there are low spots, no stock will be removed. The on-center grinding operation will not only true up the diameter size, but, it will “machine” the bar into a true round and straight part. How do you avoid these potential problems if you are not aware of the intended grinding method? Qualify, Qualify, Qualify. If “finish size” is mentioned, ask about the grinding method. And remember; “If it doesn’t clean-up, whos wallet comes out?”

-Howard Thomas, September 5th 2018

Who’s Wallet Comes Out


Between the end-user, machine shop, and/or service center, when discussing round steel shafts, there are issues with “allowance to finish”, and even with the method of grinding that will be utilized to produce the finished polished shafts. If subsequent bar finishing or grinding will be done, always let your vendor know what method of grinding will be utilized; are you centerless grinding or grinding on centers. Remember this: “When the bar doesn’t clean up, who’s wallet comes out?”

Each method will have a unique set of requirements; we will discuss those in a future note. In a perfect world there would be one semi-finished condition for all rounds. Call it Hot Rolled, Drawn, Peeled, Rough or Fine Turned. All sizes would have a standard “stock allowance for clean-up”, no matter the mill of origin, or size of bar. All lengths would also have the same perfect straightness.

Regrettably, that is just not the case. At any given time, a service center may have stock from a half dozen various bar mills. Each one has their own description of what constitutes a “Hot Rolled” finish allowance. Some mills will only give a “peeled” or rough turned finish. Another may have hot rolled, or even forged to size with allowance, not machined.

If you are selling steel, how do you come up with a textbook answer that explains which size will make the finished size? When your customer asks what size they should order to make a given part; assume they are asking: “What is the price of a car?” As a seller, can you control the machining or grinding process? Can you insure the capabilities of the operator, or even potential “movement” of the steel? Obviously, you cannot. To even attempt to help the customer, you need much more information. Qualify, qualify, qualify.

Whether you are buying or selling, make sure both parties understand each other’s needs and abilities… When the bar does not “clean-up”, who’s wallet comes out?

-Howard Thomas, August 6th 2018


Does Heat Affect Steel?

This is directed to: steel novices, steel challenged, and people who might otherwise cause harm to themselves, those around them, or pieces of steel.

So, will steel be affected by temperature? That depends. What Temperature? That depends.

Let’s define temperature as “Service Temperature”. That is, the temperature the steel will encounter where it is being used. It is worth mentioning that service temperature may be “Intermittent”, or “Constant”. If the steel is exposed to intermittent temperatures, it is not exposed long enough to thoroughly take upon itself the service temperature. (It just passes in and out of a furnace but not long enough to get as hot as the furnace.) If the steel is exposed to constant temperature, it takes-on the service temperature.

When the steel mill hardens steel to obtain specific properties, it involves heating the steel and cooling it to a very specific formula. If you are now going to expose it to temperatures that approach those used in the original recipe, you increase the chances of changing the original properties (hardness, brittleness, ductility, etc.). That is reason for caution if you are intending to do anything other than drop it or throw it.

The temperature to which the steel was originally heated were specific to the elements that were in the steel. The temperature it was cooled to, as well as the rate of cooling and even the time required to move the steel from one process to another affected the properties obtained.

Heat will affect steel based on the composition of that steel and relative to the past thermal processing that steel has undergone. 

Give or take a country mile; steels will melt around 3000°F. Whereas aluminum will melt around 1200°F. Short of those temperatures, you should not have to worry about your steel leaking off the shelf. Steels will begin to soften, however, at a wide range of temperatures based on their chemical composition and the thermal processing that got them to the current hardness.

Temperatures need not be extremely high to begin to lower the properties of the steel. Some of the very hard wear plates found in industrial applications (near diamond hard) will begin to soften at 280° to 350°F. You can cook a pork butt at 280°F.

In very general terms, if you have a very hard piece of steel that will be exposed to elevated temperatures, there is a good chance it may soften. Conversely, if you have a soft steel and expose it to elevated temperatures, you may cause hardening.

In all cases, with known grades or unknown grades of steel; when heat is involved and the steel you are using may be hardened or may be hardenable, exercise caution. (safety glasses, hard hat, gloves, etc.)

-Howard Thomas, July 5th 2018


Straightness Is Perishable

Bananas turn brown, avocados turn mushy, cars rust. Those are things we recognize as having a shelf-life. They are not permanent. They are perishable.

When discussing steel shafting, especially in the field of maintenance, straightness is an important property. If a shaft is received at the end user’s plant bent, It is not usable. You can’t grind it. You can’t machine it. You can’t install it. In fact, unless you are cutting it into little stubs for pins, or whatever, it is pretty much useless.

So, although we can all agree that straightness is important. We must understand that even if the bar has been straightened, it will not necessarily remain straightened. Straightening, and the subsequent handling, of a steel shaft is a commitment. Think of high school kids being required to carry a raw egg around for several months without breaking it. The exercise is intended to teach responsibility. It is designed to instill a sense of appreciation of the delicate nature of that item in your care.

We should think in terms of that when discussing anything about bar straightness.

Even if you require, or purchase “Pump Shaft” straightness, or, “Pump Shaft Quality (PSQ), responsibility does not end there. From the moment that product was created it began deteriorating. The severity of the deterioration will be relative to many influences. But, probably the most influential of all will be the diameter relative to the length.

A PSQ bar of 4140 Heat Treated alloy that is 3-1/2″ Dia. x 4 ft. long will be much more likely to maintain its straightened condition than will a 1-1/2″ Dia. shaft that is 16 ft. long. Then there is movement around the plant, packaging, shipping, unloading, machining, fabrication, installation, etc. It’s like those little turtles heading for the ocean once they’ve hatched. It’s a wonder any of them actually make it to adulthood.

The point is, if you are judicious, you should be able to solve most shaft problems where straightness is the rub. But know that it is not a slam dunk, just because the invoice says “PSQ”.

-Howard Thomas, May 17th 2018


What is the Length of a “Random Steel Bar”?

While there may be typical answers to that question, it is still a little like asking “What is the price of a car?” It depends on a lot of variables.

The most universally accepted random bar length would be 12ft random. A close runner-up would be 20ft random. The problem that comes into play is relative to the fact that there is no literal interpretation for random bar lengths.

Further, in the steel industry, twelve foot random may imply 10ft to 12ft random; which in reality could actually be 10ft to 13ft, or even 14ft random. If the shaft you are making has a finished length of 12ft, you would not want to order a 12ft random bar without specific clarification. Communication with your vendor goes a long way. Discuss your actual needs (“Finished Length”), with the supplier.

Perhaps, if you consider the cut-to-length price as the standard, or normal, price, then random lengths would be those lengths that are advantageous for the vendor to sell. One vendor may decide to sell 3ft, 4ft, or 6ft random bars. That allows them to utilize their end cuts. By selling “random bar lengths” they can make best utilization of their stock and pass savings incentives along to their customer.

If the customer is actually cutting the bar into short pieces, it is in their best interest to share that information with the vendor. Many times we will end up shipping 26ft bars across the country for years before we finally find out that those bars are being cut into 3″ pieces. Somehow, the total footage required to yield the number of small cut pieces was taken to be the minimum bar length. Shipping shorter pieces represented many advantages to both the end-user, and the supplier, that were unfortunately never capitalized on. Most sellers will cut a long bar in half as a courtesy to facilitate shipping; sometimes they will cut it into three equal pieces, also at no additional charge.

This minimizes potential damage in transit and often results in much lower shipping charges; not to mention potential incentive savings from purchasing end-cuts.

-Howard Thomas, April 2nd 2018

It’s Not So Obvious

“If you live in a hard-partying area of the country, you may not want to buy a new car that was assembled on a Monday. And, you may not want to shoot pool with someone whose first name is the name of a major city”. Just some considerations learned from experience.

In heavy industrial maintenance, seasoned professionals have their own hard-won cautions like the above. Those may not always be obvious. Wouldn’t it be great if those tidbits of knowledge, however, were somehow automatically transferable through the generations? However, natural powers, or gremlins, seem to insure constant attrition; constant turnover of experienced maintenance personnel, and the subsequent loss of their esoteric talents.

It’s a terrible thing to know you have solved a problem only to see things go from bad to worse because of a less-than-obvious semi-related circumstance.

Let’s say you are trouble shooting a problem where there is obvious “pitting” on the surface of a stainless shaft. For a host of reasons, pitting will eventually lead to a shaft failure. Before you begin looking into the usual suspects related to corrosion, do a little forensic investigation and see if that is really the main problem you want to solve. Pitting may not be The Big Offending Kahuna.

The shaft in question may be extra long with a small diameter (Linguini). Straightness, as in the case of a vertical mixer shaft, may be a primary concern. Let’s assume the opposite configuration of a larger diameter shaft with relatively short length (fat and stubby). In either case, straightness happens to be a key element. So, in the hierarchy of concerns; pitting is subordinate to straightness.

Most stainless-steel shaft grades, by nature of their chemistry and grain structure, retain substantial amounts of stress. Those retained stresses contribute to bow, twist, or fracture. There are grades of stainless steel, however, that respond well to thermal stress relief.  Most of the retained stress is able to be removed. (less retained stress, less movement in machining and in subsequent service). These grades resist pitting, but maybe not as well as other grades. Remember though, if the shaft never makes it into service, the potential life expectancy is irrelevant.

You know steel shafts may be straightened mechanically; so, just solve the pitting problem with a material change and then straighten the shaft. But, if the shaft configuration, or the final machined configuration, does not allow for conventional mechanical straightening, or that process would require equipment that is not readily available, or the straightener guy is just plain incompetent, experience may have opted for a steel chemistry that would be less susceptible to warp and bow; either in machining or in service. The luxury of post machining straightening was not considered an option. The best steel choice in this case may not be the one with the best Pitting Resistance Equivalency (PRE). (If the shaft never makes it into service, service life is irrelevant).

To be effective in the industrial maintenance field you must be intuitive and organized. Assuming you are, then pointing out the need to look at more than one contributor to material failure is obvious. Considering the relativity of an incompetent straightener to a pitting condition, is not so obvious.

-Howard Thomas, March 6th 2018

Does Your Stainless Remember Things?

Most likely it does, regardless of the state of your memory.

When I first heard the term “Memory”, relative to stainless steel, I was anxious to find out what it referred to. An associate with one of the stainless mills responded with this little tidbit; “Memory, regarding stainless steel, generally refers to retained stress, specifically in austenitic grades. That relates to “Movement”, or “Walking”. (Bars won’t hold straightness). Those grades of stainless (304L and 316L) have memory. They are difficult to straighten in the first place. Then, after you have manhandled them into the straightness you want, they have a tendency to return to the straightness they “remember”.

If you have a bar that looks like a tapeworm and you cold straighten it to a beautiful pump shaft, then ship it across the country, expect to find a tapeworm when you open the box.

Same with people. Take an annoying coworker. Explain why you are transferring them to your sister company. Instruct them to straighten up. Ship them across the country, and bingo!  Your sister company now has an annoying coworker.

If you want to look at this annoying tendency of memory in stainless steel a bit closer, you can start by noticing that we specifically mentioned austenitic grades of stainless. Those tend to be the most commonly used in industrial maintenance. And, of the austenitic grades two are by far the most common to the industry; type 304L and type 316L.

Coincidentally, it is just those two grades that seem to have the most profound memory issues.

They probably occupy over 70% of the grades used on a daily basis. Type 316L (we’ll talk about the “L” later) is a modified grade of 304L. It is an upgrade developed to better resist the damaging effects of corrosion. Both 304L and 316L are products that come under the general category of 18-8 stainless.

In that grouping, the first number represents chrome content, and the second represents nickel content; the two primary alloying elements in the austenitic grades mentioned.

Austenitic Stainless grades 304L and 316L;

Are non-magnetic; under most circumstances they will not attract a magnet.

Are not hardenable by thermal treatment

Can be hardened by cold work, strain hardening

Are generally of moderate strength as purchased

Are resistant to most common forms of general corrosion

Are resistant to the negative effects of service temperature to a little over 1000°F.

They possess some annoying attributes, however. In addition to memory issues, they tend to gall (Get stuck to mating parts, or, “cold weld” to mating parts), are a bit gummy, and tend to be of lower strength.


Since 304L and 316L do not respond to thermal treatment, and since the most commonly employed stress relief for steel bars is thermal conditioning. It is understandable that austenitic stainless bars retain stresses from the manufacturing process.

Since they do retain stress, and stress will not stay in a material (it will come out as movement, warp, or fracture), it is expected that those grades would have Memory; the retained stress being released as bow, twist or warp.

Once you have made pump shaft from austenitic stainless bars, you may anticipate the stress induced in the straightening process to manifest somewhere down the road. The catalyst may be: vibration, heat, torque, whatever.



MEMORY                   Has trouble with authority

GALLING                    Doesn’t play well with others


FUTURE TOPICS:                    “The One Handed Metallurgist”


-Howard Thomas, January 5th 2018