Thread: Butt Welding 1" x !/2"

I'm nowhere near skilled enough to be good enough to offer advice - only ask questions. However, a weld pulls towards the hottest part, so by welding across the top after tacking at the bottom if there is any gap the top weld contract and cause a dip. Your fit up will need to be good.
I would ask though why swap rods - a 2.5 should be able to tack well enough I would have thought.

Michael

So you are butting 2 pieces of 25x12.5mm steel together without a bevel or root gap, and welding it out with 2.5mm rods?

I'm slightly amazed that the joins actually hold

I'll leave it to the pros to give you the proper rundown, but in short you should be significantly bevelling the faces, leave a root gap and then weld up the join using multiple overlapping passes with a 3.2mm rod. If there's a dip on the top, you just run another pass and grind it back.

Guys,
Thanks for the inputs.

Michael,
The 3.2 mm tacks are heavy, they are the final attachment in that area, there is no attempt to do any running in the vertical direction.
The change to 2.5 is to avoid excessive globule development, with attendant sag, in the top corner.
The 2.5 also gives some sag, but it is possible to build up any initial holes.
The initial double heavy tacks on the outside was to ensure that any job movement was inwards (reduce the track gauge) and upwards, where it can be ground off later.
We were hoping that the 2.5 across the top would not pull the 4 heavy 3.2 tacks.

Rustarc,
The weld preparations were bevels sized 2 x 2 mm on the sides only.
Should we have more ?
At our stage of welding ability, it would be impossible to do runs vertically with our current rods.
How much root gap should we use ?

Are you suggesting we widen the gap between the rail ends, and do the multiple transverse runs horizontally ?
That we could do, and it sounds like it would be worth a test.

As a start, perhaps something like 3 mm off the ends of each bar down to within 1 mm of the bottom, in the shape of the letter 'J'.
The bottoms touching.
A 2.5 rod across the bottom, then multiple runs to the top, probably one side then the other.

I find this welding stuff quite fascinating, I am just starting to appreciate the effects of geometry and heat.

Thank you,
John.

Isn't welding up a whole track a bad idea anyway? Surely you need to factor in expansion/contraction of the tracks due to weather? Maybe some kind of bolted slip joint instead?

Legion,
Our track is virtually fully welded,
Some of it has been in operation for 50 years, and that is why we are replacing it.
The curves do move quite a bit in hot weather, but that does not seem to cause any problems.
In the 8 years that I have been observing the track, we have never had to curtail operations due to track distortion.
Maintenance is a steady requirement, but difficult to assign to expansion.

I have gotten advice from a welder acquaintance.
He suggests an increase in the bevel side preparation to 3 x 3 mm, with no bevel on top.
Use 1.6 rods.
Apply in a triangular pattern he calls a "Christmas Tree" up the outside of the rail.
Repeat Christmas tree on the inside.
Use rounded edge wheel to grind down across the rail about 3 mm.
Multiple passes to fill up to the running surface.
Grind to shape.

John.

OK.

I'm not a welder but I sometimes pretend to be on weekends. What I can say from experience is 1.6mm rods suck. Well actually they could be useful when you're welding ~1.6mm thick material but they are thin and whippy and hard to control, burn very fast and require fast reflexes to maintain proper arc length. I went through a few boxes of them, that's all, but I don't use them any more. I find ~2.5mm rods much easier on the thin stuff with the added bonus they can step up to slightly thicker material. The 1.6mm I used to cut in half to control the whipping. Otherwise they're all over the place.

Given you want to minimise heat input and your material is solid, I would investigate how it works with the biggest rods you can get and the least welding time. I'd be testing 3.2mm rods first. Easy to weld with. Big amps, fast travel.

How's your job access? That is surely a problem in situ. Can you dig holes around the join and do the sides and fill the holes afterwards? Or bend the rods, which can be very useful for getting into tight spots. What profile are the rails?

A job that thick I'd also be bevelling most of the way through and building up multiple passes as required with patience in between. Some of the experts might like to comment on pre/post-heating too, to minimise warping.

Again though, knowing the rail profile might help. I watched a video a while back, probably one of Jody's (weldingtipsandtricks) where they showed very clearly the effects of weld order and how carefully choosing it can minimise the final warp effect.

Actually the profile must just be rectangular, since it's so small. In which case I'd go with Rusty's advice.

That would minimise vertical distortion and provided you alternated sides would minimise horizontal distortion too.

With material that thick it's going to be an relatively easy introduction to vertical up welding. Job position will be the major difficulty.

You can temporarily weld another piece on top instead of clamping. Grind it off afterwards.

Practice on some scrap and see how it goes.

No pre heat it just low carbon steel.


3mm bevel per side is ok.
The christmas tree is a type of vertical up bead.
I take it that this is one rail at a time lying flat on the ground like a lizard drinking in that you are only replacing single section of rail - not the pair at one time , or are you lucky and doing pairs at a time on the bench.

If in situ, i would be doing some practice pieces first.
cheers
Grahame

My take on it is that the 25x12.5mm is sitting on it edge (lying flat would make for a rather weak rail) and the rail is welded to steel sleepers, so I imagine there's limited scope to manipulate the job.

What I don't quite get is how strong the weld needs to be.

Is the join made at a sleeper meaning both sides of the join are fully supported? That might explain why you can get away with such limited depth of weld. If the join was out in free space, I can't see 2mm of penetration around 3 sides making for a strong enough weld compared to the section being used. Even 3mm seems a bit under-par - the join is not going to be as strong as the material being joined.

For my money, the easiest welds to get right are horizontal with thick rods at high amps, so if you can do the weld with 3.2mm rods and a healthy amount of amps and lay down most of the metal in the horizontal position, that's the way to go. For vertical welds, the easiest is down, but maybe a bit hard if you can't get any upwards angle on the rod.

Trying to weld such thick steel with a 1.6mm rod sounds like a nightmare, but obviously there's no harm in trying, seeing as you have should have some old rail lying about to experiment with

Guys,
thanks for all the input, you are all giving me much to think about.

I have viewed a couple of videos and they, with your comments, are starting to make some sense.

The track is 127 mm (5") gauge with rails of 1" x 1/2" on edge.
New rails are the equivalent metric size (20 x 12 ?).
Joins are done on the ground, so approach angles are limited.
The joins are mid-way between sleepers, about 400 mm apart.
We don't seem to suffer weld drop due to traffic, but a lot of joins do have slight dips.

Practice at home is the next task.
John.

Originally Posted by electrosteam

The joins are mid-way between sleepers, about 400 mm apart.
We don't seem to suffer weld drop due to traffic, but a lot of joins do have slight dips.

If you put a straight edge on top of the join, is the dip just limited to the area of the weld, or does it extend either side?

If it's the former, it's just a matter of adding a bit more material and grinding it back, if it's the latter, then it's an issue of warpage from heat, or the weld is stretching at the bottom allowing the rail to deflect downwards.

Originally Posted by electrosteam

We don't seem to suffer weld drop due to traffic, but a lot of joins do have slight dips.

Doesn't it add to the character of travelling on a train? The dips must be significant if it is to the point of annoying passengers.

If you weld the sides rather than the top, you are much more likely to minimise warping/distortion in the vertical plane. You're looking for symmetry in weld order to use the heat from one weld to cancel another out. If the weld runs were long you would also have to worry about heat building up as you go along the weld but for 1" welds it shouldn't matter.

If you just weld from the one side (e.g. the top), warping would be an issue. It would be worth your while to investigate ways around it and learning to weld out of position. Either that or a lot of strong fixturing.

You can weld temporary pieces to your job instead of clamping and then cut/grind them off afterwards. More work but I'm always making temporary jigs out of scrap lying around and potentially stronger than clamping.

Originally Posted by electrosteam

Joins are done on the ground, so approach angles are limited.
The joins are mid-way between sleepers, about 400 mm apart.

If you're going to be doing a lot of this welding over time you might look at another weld process that's easier than vertical up stick welding against the ground. But I'd reckon you can learn vertical up stick in this position with a bit of practice.

Stumbled across that video I was thinking of:

https://www.youtube.com/watch?v=0pf-qQDslhU

It's pretty good.

There's also this old classic:

https://www.youtube.com/watch?v=2vuGlcbDwKY

which despite the age is still actually very relevant.