In Track flashbutt welding

Happy New to all,
we are looking at a program of in track flashbutt welding, cutting out rail defects, thermits and poor existing flashbutts. rail hieght varies as we have rerailed in 400 metre sections in the past. We think we know how to progress but are hoping there is a wealth of experience out there that you might like to pass on!
regards from the North West of Western Australia.

Phil

 

Colonel, we depot weld 25 metre 68 kg rails into 400 metre strings and use thermits to weld up the strings in the field. We are looking at replacing the field welds with in track flashbutts. Thermits will still be used in turnouts and for cutting out rail defects in areas already flashbutted. The axle loads are in the high 30's and we suffer rail and weld defects that lighter axle loading railroads never see

 

Gentlemen, can y'all define what you mean by the terms "Flashbutt" and "Thermit" Welding in terms that a simpleton with no mechanical engineering or hands-on railroading background can understand

 

Flashbutting is pushing a current between the ends of rails, flashing occurs and the steel at the ends of the rail get hot. As the ends flash they are pushed slowly together, metal flashes off. After a cycle of flashes a longer flash time is carried out and the rails are pushed into each other. excess weld material is sheared off, the rail allowed to cool according to its properties and the rail is ground. We flashbutt 16 rails together to make up a 400 metre string. These strings are taken out to site by a railtrain, some 450 metres long, the rail is dropped and installed into track. What we want to do now is flashbutt that string to the next string and so on. As you use, say, 25mm of rail thru the flashbutt process, we are looking at snaking the closure rails to allow for the 25 mm usage

 

Thanks Paul & Phil, I now realize that I have seen CSX crews do "Thermit" welding in the area, I just didn't know what it was called or how it was done. Even with an hour or so of setup and grinding time, it looks fairly simple and efficient.

As far as "Flashbutt" welding is concerned, that process must use an enormous(!) amount of electrical power to generate the heat required to bond two rails. My son arc welds with 1/8 inch rod and that can require more than 100 Amperes. However a 135 pound rail has a huge cross-section relative to 1/8 inch rod, so the rail must require many, many Thousands(!) of Amperes. How could you possibly generate that much power in the field? Is the rail system in NSW all electric and the track crews just reach up and clip onto the overhead wire ... Yeh, R I G H T

[ 05 January 2002: Message edited by: Hank Coolidge ]

 

In-track flashbutt welding using specialised equipment sounds like it could be a better solution to rail welding than thermit welding due to better control over the welding process. The fall down in thermit welding is the human error element where the job is rushed and/or ill-prepared causing faults that could lead to premature failure of the weld before being found through ultra-sonic testing.
I noted Queensland Rail is using Plasser APT 500 series and TG 80 equipment for intrack flashbutt welding.
Check the equipment descriptions at Plasser & Theurer's site - http://www.eurailpress.com/plasser/e/mix/mix.htm

For those interested, a .pdf document on welds with images of thermit-type set-up - http://www.eng.monash.edu.au/railway/projects/pdf/railpost-08.pdf

Gary.

 

Gary, thanks for the link to Plasser. The APT 500 units are interesting. The APT500A stated an onboard power generation capability of 400KW . This would indicate about 20,000 Amperes at a nominal 20 Volts, which I guess would do the job quite handily. I would think the feed cables would be enormous The photo of the finished and ground weld in the APT 500 series shows what appears to be a loss of temper in the opposing rail butts and in the filler material. Have you noticed this in the field, and have you heard of any premature failures in field buttweld joints?

Some of the Thermit welds on the CSX near here end up with pretty sloppy rail profiles at the butts. (I assume caused by rushing rail pre-alignment based on what you said) There obviously was an attempt to clean this up with grinders, but after a short time with heavy traffic the joints tend to become quite "noisy". Looking at the railhead polishing caused by wheels, the rails appear to be significantly out of alignment at the butts. I'm sure this is NOT a good indication for long life.

 

Spot on gents, you are all thinking in the right direction. We take 45 minute windows to complete a thermit weld, The problem from the thermits can be put down to inadequate cleanibg of the crucible and lid, poor gas pressure, poor alignment and inadequate tamping of the sleepers under the weld. With head hardened rail in track for up to a giga tonne, the rail resumes its banana shape it had when rolled. The welding crews have a tough task in aligning the plugs. We see thermits as a weaker link in our "artery' than the flashbutt, hence the push to eliminate thermits. If we go the intrack flashbutt weld way then we will be asking for a track window of around the 4 hour mark. Some are opting for welding up strings of recovered s/hand rail inthe centre of track, rerailing and moving the recovered rail up track for cutting and rewelding, so on and so on. Others are are suggesting flashbutting plugs into the rail where it lies. Hence the attempt to collect other peoples experiences.
Stay safe

Phil

 

As I understand, the Flashbutt weld's heat is generated by microwave passing through water-cooled lines similar to the way we quick-melt steel in a foundry. Most weld failures were from improper quenching of alloy rails after the weld is made.

Some failures were contaminated with silica imperfections from rust, grit, oil and rocks falling into the crucible, or trapped between rail heads. Thermit (That was invented to burn holes in tanks, and cut bridge supports during World War II), sometimes crystallized the parent rails from over-temperature. Slow cooling was found to releave temper strain similar to annealing, but presented a small area of relatively softened rail right at the weld. Re-heating, or "Drawing" seemed to solve some failures, by leaving the tough foot and web, while re-heat treating the harder head surface. Without this step, fractures were found to occure within a shorter time from wheel pounding, since the usual weld left the metal brittle. It was suggested that welds be made only over sleepers, as an added support to prevent cantilevering stress that would be presented when the weld was hanging free between sleepers. The idea worked in theory, but was rejected at the time, due to the cost of inserting and tamping an odd sleeper. In any event, if either type weld fails, the result will be catastrofic!

 

Have any of you noticed the following?

The CSX rails in our area are developing fractures at the welded joints. The fractures consist of a horizontal flaking across the rail head, up to 4 inches along the rail either side of the weld. All cases appear to be at factory welded joints which I assume are "Flash Butt" welds. The flaking can go to a depth of 1/16", then it smooths over with time and wheel hammering. Then the next layer down begins to flake, and the process starts all over again. The process takes about 2 years from start of fracturing to full smoothing. Traffic is about 30 freights a day, every day, running at 50 mph. Power normally consists of 3-4 4000-6000HP GE's or EMD's, the trains are long and heavy.

I haven't seen any of the field "thermit" welds fracture.

 

If you have ever watched wheels as they roll over any one spot while the train passes, you have noticed the rails "give" a little under each wheel.

The rolling wheel imparts a "work hardening" much like case hardening to the rail head surface. This "case" will become harder and less malleable than the rest of the rail. Case hardening sometimes goes .080" deep, where in heat treating a "case" of hardening the steel is only a skin at the surface. It can also be forged into the piece by repeated pounding usually in a trip hammer, where it can be made to harden a surface up to 1/8" deep. It depends upon the alloy.

Thus when the rail is bowed down, tensile stress down at the rail's foot tries to stretch and the rail head's wheel surface tries to compress. This compression can occur with less strain in the malleable rail body and foot, than in the harder more brittle "case" area which causes the "case hardened" area to fracture allowing it to peal away from the rail body. When this cased chip finally fails and falls away, the succeeding pounding of each wheel, begins the process all over on the "open" surface below.

This quite often happens when forging a billet and scale (totally burnt iron) gets into the trip hammer and becomes imbedded in the billot. If not detected, it becomes a cause of failure when under stress and strain later.

In the case of thermit welding, scale or slag is a danger difficult to prevent. Usually an argon gas is introduced to prevent excess oxygen from burning the rails like a cutting torch, yet allow the rail butts to actually melt or at least to become plastic so they can be fused together. Any particles such as rocks, scale, slag or cinders become trapped surrounded by the molten steel. The void they cause weakens the weld.

That is why we X-Ray every weld and forging used by the Navy and aircraft assemblies.

 

Nice explanation. Thanks Wayne

I guess the reason I haven't seen railhead scaling with in-field Thermit welds is that the immediate weld area has not been case-hardened and is still malleable. However, with factory welding, all weld joints are case-hardened as part of the Flashbutt process.

[ 09 January 2002: Message edited by: Hank Coolidge ]