Mr. Mizer,

I was not suggesting a new technology but one that is sixty years old and has been applied to tens of thousands of locomotives worldwide. While I was making a somewhat less than serious comment to M. Austin, there is an interesting problem brought up by your comment.

What denotes an established railroad practice? Where is the boundary of history, and where does it become too modern in a world that has never stopped using steam and building new locomotives to new designs?

Having researched UP’s drawings and practices extensively, I see no clear divisions. Practices and designs evolved continuously, and that is exemplified by references on sheets “forward” and “back” in UP’s material. This did not stop in the 50’s as I have seen drawings for modifications to steam locomotives from the 70’s, and improvements are still occurring today. The fact is that we are still learning to do things better. We are seeing things in other industries and in other parts of the world that we can use to increase our locomotives’ performance and increase their reliability. It is as important to me to continue practicing innovation as much as preserving the past as this aspect is what has made the steam locomotive competitive all of its existence. Today, the competition is not with the diesel, but with other forms of education and entertainment sustained by disposable income, and we need all breaks we can get.

We would be happy to help you find solutions “outside the box” in addition to finding something relevant to a particular time period or place.

Matt Janssen, Foreman - WRC

The location of the above-mentioned documents (fillet welded staybolts) and others is files.wasatch-rr-contractors.com. Adobe Acrobat Reader version 6 or later is recommended to view the documents. If you have any problems opening the documents, please do not hesitate to notify us.

Dear Mr. Janssen

You wrote "Though this may be true, what I have learned from research is that bending stresses and not tension stresses are the primary cause of breakage in normal service. "

Could you explain to me how tension loading is not found in a bending load in at least a part of the material loaded? Furthermore what is the effect of essentially putting a crack into the sheet staybolt interface (water side) with respect to various systems of corrosion?

Respectfully,
Robby Peartree[/quote]

Bending sounds right to me - based on my experience. I reckon a tension exceeded staybolt would stretch first, then seperate somewhere towards the center of the bolt. I have NEVER seen this! The only broken stays I have seen and repaired (or not) have broken very close to the outside sheet. Given it is usually thicker and cooler than the inside sheet, and therefore less prone to giving a bit as the stays move under load, bend fatigue stress coming at the weakest point in the thread root nearby seems logical to me.

Did I explain this clearly enough to have really smart guys like Matt tell me where I am mistaken, and why?

dave

Dear Dave.

In a true bending you have areas of compression forces and areas of tension forces based upon what line (at zero load) within the material is node (zero force). Anything with wider radii than the node line is under tension. Everything with smaller radii is under compression. Material failure modes change depending upon the material and how a force is applied, so the description of forces would be critical in predicting a possible modes. I hope this draws a mental picture of what I am describing.

Respectfully,
Robby Peartree

Mr. Peartree,

You must have figured out I am not an engineer from my comments. I follow your line of thinking that when you divide the cross-section of the staybolt you get zones of local compression and tension and that it is these forces going into to the material’s plastic limits that causes fatigue and breakage. Dr. Tross looked at the direction of these forces which is exemplified by cracks in the tension areas of the bolt. He found that the cracks were usually on top of the bolt on the wrapper side and on the bottom of the bolt on the fireside. Furthermore, as you move away from the center of the sheets the clock positions of the cracks turn to point to their center. He explains that the firebox sheets operate in the plastic range under high heat transfer conditions and that the sheets gather and deform somewhat as the hotter firebox tries to expand relative to the wrapper. When the fire is killed and the firebox sheets cool into the elastic range they bend the staybolts as the temperatures of the wrapper and firebox continue to equalize. The effect is most prominent in the hottest areas of the sheets, and we refer to its appearance as corrugation. The shrinkage of the firebox also causes the wrapper to “hog out” like an old wooden ship.

Dr. Tross gets into bending lines of staybolts derived from the Mohrishe Skylift Principle to describe bending forces in his paper, “MATHEMATICAL-GRAPHIC METHOD FOR THE CALCULATION OF RIGID STAYBOLTS WITH NEW DESIGN SHAFTS UNDER ALL POSSIBLE DEMANDS AND CONDITIONS.” The math is a little beyond me, but the basic idea is to calculate bending forces at fifteen cross sections of the staybolt. He found that bending forces are highest closest to the sheets. Through the method in the paper, he shows how to derive an ideal diameter at these cross sections to distribute the bending forces more evenly. The bending across the staybolt body becomes a smooth S- shape and internal stresses are kept lower and relatively equal for a given bending load from the firebox sheet. This allows the staybolt to deal with higher total bending stresses before the material at any one cross section plastically deforms and fails. He later shows some generalized proportions practical to manufacture that achieve dramatic improvements over cylindrical shaft staybolts.

For me, it is easier to describe to people that bending breaks staybolts, however, I believe you are absolutely correct.

S - shaped staybolts? Hmmmmmmmmm............seen this done once before by a certain individual in a very small New England state. Hope you aren't recommending that.

Gents, I think we all understand that "bending" is shoving compression towards one side and tension towards another at a given part of the staybolt. I was contrasting that in my mind to a staybolt being pulled straight out towards both ends, without any sidewaysness or torquiositude of force.

So are we all agreed they break by bending causing concentrations of force at critical locations close to one sheet?

dave

Dave,

It may be uncomfortable to think about, but quite a bit of elastic (temporary) deformation occurs to a boiler in operation and also plastic deformation (permanent) occurs over time in normal service. These stresses can be reduced somewhat through improved boiler design, combustion system design, fire-up practices, and washout practices, but they still exist.

We are agreed about the breakage areas. You have to remember that bending shapes are very minute before the stresses cause breakage. If you were to look at the bending that occurred to cause failure next to the sheet and magnified it, the middle of staybolt shaft remained rather straight and the bending was concentrated just inside the sheets almost at a point (or bending line.) Little of the stresses were transferred to the rest of the staybolt body. The Tross method manipulates the profile of the staybolt into a parabolic shape (radius) with an increased diameter at the traditional breakage area. By matching the diameter to the stress, bending lines become almost undefined and equal bending occurs across the staybolt in an S-shape. This allows the staybolt to remain elastic (returning to straight) for higher stresses than a cylindrical staybolt.

I urge you to read the papers on the WRC website. This is just a scientific approach to the “reduced shaft” staybolts that is more effective. It will all make sense after a few reads.

Matt

The failures of staybolts are caused by many factors. Staybolts in regular service operate in a complex environment and each failure must be examined as an individual incident. The function of a staybolt is to give structural integrity to a given area of the pressure vessel. Different designs of pressure vessels create different advantages and disadvantages of design. These differences of design create different issues that need to be considered with each one. Therefore depending upon where the bolt is in the boiler would result in different factors causing its failure. Another issue to consider is the when new research is undertaken, the investigators biases and beliefs will have a certain effect no matter how objective they are. Further the material quality control has evolved over the years as Metallurgy has grown. Many steels from the 1940’s and 1950’s would today not even make good scrap due to their tramp element contamination. So when working on older machines a number of factors can influence material behavior.

let us consider a 1020 steel fast with a yield point (YP, the pressure above which permanent deformation occurs) of 50,250 psi. and an UTS of 64,000 psi. According to CFR 49 Part 230.24 the allowable stress on any component shall not exceed ¼ of the UTS (Ultimate tensile strength) of the material. In most cases ¼ of the UTS is well within the elastic range of the material. The maximum stress allowed by 49 CFR part 230.24 would be 16,000 psi. So for me to achieve a level of stress to cause permanent deformation of the material I would have to find an unaccounted for pressure of over 34,250 psi to cause permanent deformation. If Permanent deformation takes place, then stresses were greater than ¼ the UTS at some point. If our models for the calculation of stresses are not that far off so what causes the change? One possible change is a change in the value of UTS and the YP are temperature dependent. These values change dramatically when the metal passes thru its ductile/brittle transition temperature which happens on many steam locomotives at a temperature between freezing and room temperature. Subjecting a staybolt to pressure below the transition temperature can have devastating effects to the staybolt. In fact a few hydrostatic tests have come to a sudden and exciting conclusion due to a brittle failure during the test. This means that the environment is changing the structural conditions of the boiler. This could be one of the reasons of having a lower maximum stresses on stays and braces found in 49 CFR part 230.25.

Boilers are corrosive environments. At higher temperatures chemical reactivity typically goes up thus bringing its influence into the life of the bolt. Chemical attack will preferentially occur where the bolts are under the most stress. With corrosion, a key influence on where it occurs is the shape of the surface. By having a change in the metals surface we create the condition by which corrosion will preferentially occur. A particular concern is a stress concentrator shape such as a 90 degree angle. Unfortunately, for this method of applying staybolts, one of those locations is at the end where the bolt and side sheet interface. With this welding method we create a concentrated stress zone within the weld that is still potentially subject to corrosion. By doing a full penetration weld or the older mechanical methods of applying staybolts we eliminate a lot of potential issues within the joint The issue of heat effective zones from welding can be addressed by a pre heat/post heat program to the material. In a lot of welding process in industry you have set procedures to cover metal prep.

Often in our efforts to model a given environment we simplify the model environment due to a number of factors. In this effort to model the environment under study we sometime lose sight of what were trying to do. I have heard plenty of projects that were well meaning but did not understand the basic characteristics of the process they were trying to improve. Without that knowledge we can do more damage than good in spite of our best efforts.

Respectfully,
Robby Peartree

Mr. Peartree,

Well said. This level of thinking is incorporated in to the current German TRD (Technical Rules for Steam Boilers.) In the end, they still endorse fillet-welded staybolts. Perhaps wrongly, I don't know. I can tell you from having been there that an entire firebox application of full-pen bolts is impractical and dangerous with extensive preheat to the level of mitigating initial staybolt stresses. I can also tell you that tens of thousands of boilers have been built and repaired with fillet-welded staybolts (and still are with modern materials) with minimal problems. Check out the A1 project. Draw your own conclusions. While I do not disagree with you in laboratory conditions, conditions in reality are such that it is more likely to achieve reliable attachment with the fillet weld. Materials are also less than ideal. Though that does not justify them, it is reality. Then there is the issue of economy. Though it is “wrong” to address this in code meetings, too much impracticality kills.

I urge you to attend the next ESC meeting or prepare a paper with your thoughts as I feel they are important for the committee’s consideration.

Regards,

Matt

Initially I was very encouraged about the possible use of fillet welded stays. I believe they can be installed more reliably than with full penetration welds, and certainly there is less prep work involved. From an economical standpoint, fillet welding a straight bar is the quickest method and should work reliably for a period of time.

Making special bolts with a tapered configuration of some sort reduces the economic advantage. Application of the bolts in a “neutral” manner with the fillet weld closer to the center of the hole further reduces the economic advantage. While it has been shown that both these methods probably mitigate stress failures as much as possible, I don’t think the economics justify the application in steam locomotives used in tourist service. There may be some justification in class 1 service, but I find it hard to justify in the preservation world. I can state with some confidence that it is less expensive to install threaded bolts in the traditional manner than it is to make specially shaped bolts, where length is critical and attach them in the neutral manner described above. Given that our #90 has over 400,000 miles on it since Great Western days, and there are many bolts that have been in service the entire time, the threaded staybolt has some considerable merit.

Probably my greatest concern is about what I believe is called crevice corrosion. I had the opportunity to see a bolt that had been fillet welded in, I believe the 1960’s that failed because of this phenomenon. It is my belief that, if there is a future for steam locomotives, it is in tourist service. If there is a resurgence of steam on the class 1’s, those locomotives will probably not even resemble what we know now. That being said, we are repairing, rebuilding, and building our historical boilers for a long and not necessarily severe service life as opposed to intense service over a definable lifespan. I think this is where the corrosion issue comes to the surface when we think that this boiler may be around 75 years from now, as are many of our existing boilers.

It seems to me at this time, that reinventing the wheel may not be in our best interest. Threaded staybolts, in my opinion, seem to be the best option for boilers in tourist service. They have proven themselves, we know their weaknesses, and they are relatively easy to install without resorting to the “W” word. We can predict with some certainty what condition threaded staybolts will be in 50 years from now. I do not think we can predict as reliably what condition full pen or fillet welded bolts will be in. We could be making a booby trap if we aren’t careful.

For the record, I do not consider this matter closed, but more study is needed, with attention paid to a service that was not necessarily considered in the staybolt tests I have seen to date.

Good points - I'd be surprised if there isn't about 60 years of documented history in Germany (knowing their cultural propensity for bureaucracy and technical excellence) relating to reliability and longevity in service we can look at for fillet welded stays.

I must take issue with the statement the full penetration welded stays are dangerous, however. Many boilers have been built with that technique and worked in successful service for many years - Jupiter and 119 come to mind.

dave

I just thought of a problem I had with threaded stays several years ago. The locomotive concerned had been retrofit with piston valves not particularly well, and about half the steam blew right past the valve rings out the stack for decades of use in service. Once the valve problems were solved, the problem quickly became the rigid stays, which were incapable of staying caulked for more than a couple weeks.

After trying everything that seemed reasonable, I cut out the plate around one bad stay and discovered that the stay was good, the firebox was good, but the threads connecting the two had mostly disappeared. I reckon this was caused by decades of force firing to try to get some use out of her.

Repairs were made with full penetration welded rigid stays. None opened up again during the few years I was still involved.

So, I wonder: isn't the thread just a long spiral crevice for corrosion to infect? I know it reduces the effective diameter from nominal to root, and can set up a stress riser if threads are cut rather than rolled.

I wish I could open those files Matt refers me to.

dave

Dave,
If necessary, I will generate a paper copy and mail it to you. Please let me know if this is what you need and your address.
Thank you for your interest.
Dave Griner

Dear Matt Janssen,

Could you enlighten me and others on the value of bringing such a paper before the Engineering Standards Committee? While I know of its existence most of its actions are not apparent. Most boilers in RR service are under either FRA or state Jurisdiction and what role is there for the Engineering Standards Committee?

Respectfully,
Robby Peartree