Dear Mr. Linn W. Moedinger

Crevice corrosion is a definite possibility with this type of application. With the Fillet weld on SP 3420 front flue sheet we have tested as much of the weld as possible. Because of the limitations of dye penentrant we went with dye mag particle testing. While the weld where we can reach on the Smoke bow side shows no signs of cracking we have other concerns. The first course (3/4” thickness) has a bulge where the weld was made. This budge seems to vary over the circumference of the coarse which lead to thoughts of uneven cooling rates. The testing by a Natural gas pipe line company confirmed our suspicions as they stated that they saw signs that the weld had cooled rapidly. Fortunately field metallography and diamond hardness testing done on the First coarse weld and tube sheet showed that the strength of the materials were close to what we had expected for steels employed in this type of service. Unfortunately because of our inability to reach the water section weld due to tubes and the superheater block on the smoke box side were still left with a lot of potential questions. Moving forward technically (aside from political issues) has been difficult partly because of our lack of knowledge of the procedure and the unknown standards used. That is why I was very interested to hear Gary Bensmen mention to me in 2002 of an N&W standard for this type of repair, 12 years after we began discussions of this very issue with his then employee Robert Franzen.

Having time looking at failure analysis samples of things that worked in principle and not in practice I can easily state that I am very confident in the performance abilities of a full penetration weld done correctly over a fillet weld. We have already exceeded initial the design life of these machines. We need to carefully consider the future results of our actions today.

For those who thing were some how sheltered legally consider this. Over the past few years I have spent a lot of time on the phone talking to people dealing with the repercussions of the natural gas pipeline accident near Carlsbad, NM which had tragic results. For many of them the hard part was not dealing with the NTSB and their findings but rather the time in Albuquerque in Federal Court. It is a place that I hope no other has to experience Extra time and effort in maintaining our beloved machines is always a strong deterrent toward spending time with lawyers representing others who now see us as the financial solutions to their problems. And if you are really not lucky the lawyer you face will be bringing criminal charges against you. This type of legal action could shut down this industry. Careful and deliberate decisions must be made on each machine that we approach.

Respectfully
Robby Peartree

Mr. Peartree,

The ESC is the closest organization we have in the U.S. today to the intellectual societies who dealt with this technology in the “days of steam.” It wrote a great deal of the FRA and NBIC code currently applicable to the steam locomotive today, and will likely be the prominent specialty influence for these regulatory bodies in the future. Meetings are open to the public, and all members are receptive to the opportunity to learn and discuss issues.

Matt

I must weigh in on the statement about full penetration welds. I agree that a full penetration welded staybolt applied according to the sketches in the code is probably the best attachment. If you combine that attachment with a Tross style bolt you may be close to perfection. This application would have a perfect fillet at the root of the weld with absolutely no undercut on the bolt. The bolt would have a larger diameter thru the sheet and then taper in the most mechanically advantageous way to the longitudinal center of the bolt and then mirror itself in the second sheet. Post weld heat treatment would be performed in such a way as to eliminate all stresses from fabrication. No backup washer would be used for the stay weld as this invites crevice corrosion. All welds would be radiographically examined to ensure integrity. Do this in the entire firebox and you have something to write home about.

You’re a welder. It’s 3:15 pm. You have been lying on your back, standing on your head, welding between your legs, and your feet, butt, and elbows are asleep. You have been welding staybolts all day and you have 45 minutes to go before you can have a beer. Oh yes, by the way, you are also a human being. You are not Titanium Man. Even if I allow that all the welds made in the previous 7 hours and 15 minutes are perfect, I have a bottle of JD Single Barrel that says the next 45 minutes isn’t going to see welds from heaven. And then of course there is verification. Who is the little guy with the x-ray camera and film that is going to take the snapshots of all these welds?

The code calls for these perfect welds but requires no verification and provides no realistic quantification of the process other than having Titanium Man pass a welder’s qualification test. If the thing blows up, the code jockeys can say, “Oh well, you didn’t do it right, you’re going to jail.” Nice system.

That being said, having welded for 35 years, I can guarantee that bone tired, thirsty, and half blind, you can make a better fillet weld than you can an open root full pen weld. To me that is the issue at hand, as well as keeping the heat affected zone out of the breakage zone. A fillet welded bolt that is snug in the hole will tend to bend on the water side of the sheet. It will induce stresses in the sheet, but they should not be any different than a threaded stay. The problem seems to arise with assembly induced stresses because of welding on one side of the sheet. This is where the full pen weld is superior, or the neutral fillet weld done nearer the longitudinal center of the bolt hole.

At this point I am thinking that to minimize crevice corrosion, or at least have it be no different than it would be in a threaded bolt, the bolt must be snug in the hole. I currently have no practical solution to the stresses induced by welding that seem to be mitigated by slightly enlarging the hole, which makes a crevice.

My main issue with full pen welding is that it is required to be done perfectly, and I have never, ever seen one done that way. Obviously a less than perfect weld works – that is all that is out there, and some I have removed are way, way less than perfect, but they held. Mr. Peartree talked about liability. Every time you do a full pen weld, it is a liability under the current rules. At this point, I don’t have any idea how you would quantify a less than perfect weld, and to me that is a problem. For that matter the only way to verify that a weld is perfect is to send the little inspector in with the x-ray machine, or rely on the perfection of Titanium Man – the super welder.

The logic path I am forced to follow at the moment is:
1. Threaded Staybolts
a. In service for over a century
i. Heavy steam service
ii. Intermittent tourist service
b. Proven technology for this service
c. No structural welding
d. Mechanical attachment
e. Quantifiable installation (did the threads survive going thru the sheet?)
2. Full Penetration Staybolts
a. In service around a half a century
i. No heavy steam service
ii. Moderate steam service (maybe 20 years)
iii. Intermittent tourist service
b. Not proven in steam service to the degree of threaded bolts
c. Structural welding
i. Unrealistic standard
ii. Not quantifiable
d. No mechanical attachment
3. Fillet Welded Staybolts
a. In service maybe 70 years
i. Heavy steam service
ii. Intermittent tourist service
b. Proven in steam service
i. I have seen no data on 70 year old fillet welded boilers
ii. Appears to work well over the short term, 25 – 50 years
c. Structural welding
i. Other than scheduled removal, condition is not quantifiable
ii. Stresses induced by one side welding
1. Questionable cost effectiveness of more elaborate design to mitigate construction stresses
d. No mechanical attachment
e. Crevice corrosion may be an issue

This little logic path may be too simple and there may be answers to some of the concerns, I just haven’t seem them. This is a very enjoyable thread and maybe we will find the answer to be 43 until all is said and done.

Excuse me, I may not be entirely correct. I know that the ESC along with the FRA made sense of the requirements of the CFR part 230 as it applies to these later days of steam but I did not hear that the ESC helped rewrite NBIC code. Is this correct and if so what were the relevant parts of the NBIC code that the ESC rewrote.

Perhaps someone from the NBIC could comment on this. I subscribe to the NBIC and I did not receive any addendums that appeared to match the CFR part 230. Perhaps I should go back and review my NBIC code book.

The ESC wrote Appendix 3 for repairs to steam locomotives and is responsible for maintaining it as an NBIC sub group serving at the pleasure of National Board. The appendix was written before the work on the CFR. It was originally written as a repair code only with part 230 of the CFR being the inspection code. Appendix 3 now has inspection guidelines as well.

Dear Matt Janssen,

What can you tell me of the book” La Locomotive a Vapeur”?

Respectfully
Robby Peartree

Mr. Peartree,

"La Locomotive à Vapeur" was written by André Chapelon, who worked as a locomotive designer for various French railroads throughout his career. It is the view of many of us that he is the one who started to bring thermodynamics and scientific study into locomotive design replacing rules-of-thumb and ignorant factors of safety that had prevailed in design previously. He also studied American design very thoroughly, and had great respect for the mechanical reliability of our locomotives. His later work represented a culmination of the best features for an international perspective. L. D. Porta studied Chapelon’s work and continued development of the steam locomotive by furthering the scope of the study of locomotive design (such as German boilers and Tross) and also began to incorporate internal combustion engine design elements that could benefit the steam locomotive (such as diesel piston rings & materials.)

The first edition of "La Locomotive à Vapeur" (1938) covers Chapelon’s theories of locomotive design. It is in French only, and being very rare, will cost $1500 or more if you can find it. I would love to have one. The second edition (1952) covers his international survey of locomotive design but covers less of his own theories and is overall less technical. Both were published by Baillère in Paris. Camden Minature Steam Services translated the second edition to English recently, and copies are still available.

Matt

Hey Matt, are you aware of any English version of Porta's cylinder tribology paper, or any other work in which he details his methods for cylinder liners, pistons and rings? I have read all i can find, without locating detailed descriptions, and my ignorance of Spanish is truly appalling - just some Cuban profanities from growing up in Miami decades ago.

dave

Dear Mr. Moedinger,

The science behind full penetration welding or threaded bolts shows advantages for both but the results is you have a safe repair when done properly. Unfortunately the same can not be said of fillet welds for this application. You’re creating a crack that reduces sheet thickness for a given length. While welds are typically stronger than their surrounding material it will not overcome the significant reduction in cross sectional area of the effected sheets. This opens the possibility (via a number of possible causes) of rupturing sheets instead of just breaking bolts.

One facet of engineering today is predicting your failure mode for a given machine or component. By either determining or controlling the failure mode you can manipulate the failures impact on the environment which should help control your liability. By controlling the fracture to the volume of the bolt within the pressure vessel we can effectively eliminate some forms of catastrophic failures. One of the advantages of steam era shops was people were given one task that they did repetitively. This repetition created great skill in a particular task. In the world of today, steam locomotives are repaired by people who try to know a lot about a lot of different aspects of the machines. Today we do not have the economic ability to afford people who have just one narrow skill base. Yet the science that determines how our beloved machines behave does not change based upon our economic conditions. So in planning our maintenance it beehives us to consider both the scientific and human factors in all of their aspects.

In this day and age of the over production of lawyers that need their yearly new car we must consider the potential results of our actions if we wish to minimize our contact with the legal systems with the title of defendant.

Respectfully,
Robby Peartree

Mr. Peartree,

I follow much of what you say, but I don't quite understand what you mean by creating a crack or thinning the sheet and risking a sheet failure.

As I see it, if you put a straight bolt through the sheet with a snug fit and put a 3/8" leg fillet weld around the bolt, the overall sheet strength is probably stronger than with a threaded bolt since you don't have an unreinforced hole where the bolt is. The weld and the bolt serve as compensation for the material removed from the hole. Obviously, if done correctly the efficiency problem encountered in some roof sheets goes completely away with a full pen weld since the efficiency is 1. The efficiency would probably be less than 1 on a fillet welded bolt, but not as low as on a threaded bolt.

The three problems I see in fillet welded bolts are: 1. Creating a crevice for corrosion. 2. Distortion of the sheet with the ensuing stresses because of structural welding on one side only. 3. Verification in the long term of the structural integrity of the fillet weld. The first problem could be mitigated by having the bolt fit snug in the hole. The problem could be eliminated by only having the bolt enter the water side of the sheet by a tiny bit, say 1/32", and fillet welding on the head of the bolt and inside the hole. This would almost eliminate the crevice and be no different structurally than the Tross neutral welded (my terminology) bolts. This would take care of the second problem as well. This would be time consuming because of the precise control of bolt length that would be needed and welding on the head of the bolt goes against all conventional wisdom - right or wrong. Other than periodic scheduled replacement, I don't see a solution to the third problem.

I may be missing something in your post or possibly I have misread it in some way, but I would be interested if you could further explain what you mean by creating a crack that reduces sheet thickness over a given length. Sometimes it takes a couple tries to pound something into a thick-headed dutchman.

Dear Mr. Moedinger


I am using the term crack to illustrate how I see the metal bridging the crevice will take the load microscopically. While metals are able to take a given load, loading over certain physical characteristics creates certain internal forces that can create localized deformation and fracture. As the effects of time take their toll you see progressively more and more damage at the crevice or crack tip.

When looking at the loading one must ask just how thick will the weld be? The Czechoslovak State Railways application for 4-8-2 wheel arrangement locomotive boiler called for a 6 mm (thick weld (app 15/64”) over the 1 mm (1/32”) gap between the edge of the stay bolt and the sheet. That means if your side sheets are 3/8” thick you loose 37% of your cross sectional area of the sheet at the point where the “crevice” is bridged. Each crevice has an approximate area of 0.050 sq. in. If you have 3000 staybolts you are looking at a sum of 150 sq. in. of area that is effectively 15/64 of an inch in thickness before corrosion or other factors begin to affect the joint. After a few years of service you will be looking at repairing staybolts due to the metal thinning at the joint. If this does not happen the line of failure will not be a bolt breaking across its diameter but thru the weld between the sheet and the staybolt.

There are other questions one must ask. Will the staybolt be centered in the hole or will it sit at a certain offset? Do you have one area around the bolt with the 1/32” gap and another area closed? What is the water chemistry that the boiler will be subjected to in service. Material behavior is driven by a number of factors that are not always obvious. While the mechanical design of steam locomotives has not marched with the same urgency that it did when steam ruled the rails thru out the world there have still been scientific discoveries (particularly in metallurgy) that can help us both explain and improve the behavior of our machines today that were not known fifty years ago. We have to be careful of implementing new technology because unintended consequences have dramatic ways of raising their ugly head. The other thing that should be considered in engineering is just because the design works for one discipline of engineering it may have issues in another. While fillet welded stay bolts are cheaper and faster to install on a boiler than a full penetration weld they are not as safe. Today we know a lot more about the way the metal will behave than they did in the 1920s and 1930s. It behooves us to understand how a staybolt replacement technique behaves over the life of the vessel vs. short term benefits. These machines have outlived there design lives. We have to consider a lot of issues today that they may or may not have considered in the original design.

Robby Peartree