Welding 4130 to other unknown steel (either 4130 or 4140)

[leif20]

I am part of a student racing team team at Western University (London Ontario). Each year we design, build, and race a small formula-style open wheeled race car. In previous years the entire car was a 4130 profile-cut steel tube space frame, which I helped to weld. In 2015 we switched to a hybrid chassis design - the front is a carbon-fiber sandwich panel monocoque, and the rear is still a cromoly space frame. We use a motorcycle engine (Honda CBR600F4i) mounted to the space frame with 8 separate mounts, 4 on either side. The mounts are tophat style slugs placed into the cromoly tubing. Typically we use a mild steel to make the tophats and then weld the parts to the 4130 tubing using a soft filler such as ER70s-2.
This year one of our experienced team members made the tophats out of a peice of stock he found on our material rack. He complained of some of the parts being very hard to part off on the lathe, and the stock showed some really wavy marks where the parting tool was. The material had a very shiny finish to it (i will post some pics shortly) and didn’t look like regular mild steel. I took the stock to a machinist who told me it is not regular steel, it is definitely something harder like 4130 or 4140.
The same guy welded these parts to the cromoly tubing to complete the engine mounts. He used unibraze s-6 (ER70s-6) MiG wire as filler since the gaps were very small. I know he did not preheat the tubes or the tophats before welding. These are photos of the finished welds:





You can see some slag on the weld beads. I think this is relatively normal. I am more concerned about the possibility of these mounts cracking. They are very highly loaded in the car (they hold the engine in place, and the engine drives the wheels via chain drive) and are exposed to large amounts of vibration as well. I know in my experience welding hard materials (without preheat) is that welds can crack. I am very concerned that these welds will crack and cause our engine to start rattling around. Usually we use mild steel for the tophat material and weld it with ER70s-2 as the filler and we have never had an engine mount failure. This year, with our unknown tophat material (could be 4130 or 4140) I think there is a significant chance of failure.

I have a couple of questions:
1. Knowing that the tophat material is not mild steel and is probably a higher carbon steel, would you be concerned with this weld failing? This is a highly loaded area experiencing lots of vibration.
2. If we need to redo the weld, how far down the 4130 tube should we cut to ensure that there is no remaining material from the original weld?
3. If we redo the weld with ER70s-2 and mild steel top hats will it be more likely to crack since we have alredy put a heat cycle through the cro-moly?

I appreciate you making it this far and any help you can offer. I have some experience welding and fabricating but I would rather not learn this lesson the hard way. This is our car on the line and the stakes are high!

[MMW]

I am far from an expert on this but if it was mine or a customers then I would redo them. That is not to say that they will fail but rather the unknown of if they will hold up or not. You guys are building a complete car so why let a small item pass that you are unsure of?

Assuming all the tubes are about like you pictured I would suggest cutting new tubes & top hats to weld together. If for some reason you need to use the old tubes then you need to go past the haz for peace of mind.

As far as the weld itself it looks to me like inadequate gas coverage or it was welded to hot.

[ryanjones2150]

I've tried responding here for about 30 minutes and keep getting some stupid error message...sorry. I'll try again later.

[H80N]

From the pics.... those welds look COOKED to me.... not enough amperage.. moving too slow = too much heat input ..while slow to puddle

Between that and the unknown material / filler..... too many unknowns for my comfort

I would junk those parts and remake them... the safety issues are too great.... do not attempt to rework or "fix" them

If you use 4130 on the new parts ER70-S2 is the recommended TIG filler

Just my 2cents worth ...

[griff01]

These welds sorta look like they were done with silicon bronze instead of ER70. Also, too hot and not that good either.

Griff

[pennswoodsed]

The value of the time spent looking, considering the safety of these, exceeds the cost of new material and the lesson learned.

[leif20]

Hi everyone,
Thanks very much for all the replies. I've gotten bogged down with work so I haven't been able to respond.

On my end, I spoke to a number of professors at my school and also got a few conflicting answers. However, one discussion I had with a materials prof allowed me to finally get some sleep and feel better about the situation. We discussed TTT diagrams and how the cooling rate affects the microstructure of the metal after welding. The TTT diagram shows the temperature of the material on the y-axis, the time to cool on the x-axis, and the phase transitions as functions of slope. A plot for 4130 is shown below:



I have placed the red marker on the y-axis indicating around where the temperature of the material after the weld puddle cools. Based on experience, it takes the welded area over a minute, generally around 3 minutes to cool to room temperature which is the bottom of the graph. The areas marked with Ms and Mf are the points where martensite forms in the material, and since martensite is the hard and brittle phase of iron it is likely this that causes the brittle fracture. This also explains why it is important to pre and post-heat welds. If the area where you are welding is right next to a cold area (think a block that wasnt pre-heated and most of it is at room temperature), this area will have a very large temperature gradient and the local rate of cooling will be much greater there. The weld would be most likely to fail in that spot.

Compare the 4130 TTT diagram with the diagram of 4340:



This diagram shows that 4340 has phase transitions that occur much later than the 4130, meaning that Martensite formation will occur sooner with a faster rate of cooling, and Austenite or pearlite formation requires slower cooling rates to form. This is why 4340 is regarded as a poorly weldable steel as it requires more attention to proper heat treatment. We have reason to believe that the steel we used for the engine mounts is NOT 4340 as we have a labelled piece of stock in the shop and it appears much different (outer surface brighter, machine surfaces shinier... very accurate I know). Therefore since the material used is either the same as the chassis tubing (4130) or slightly different (4140) I feel much more confident in the parts.

Before speaking to my professor I did not really understand why metals can become brittle during the weld process, I had only seen it myself when welding on tool steels. In the past when TiG welding tool steel I had not preheated the material and the welds cracked almost immediately. The TTT diagram explanation allowed me to get a better understanding of the microstructure transitions and put some numbers to all of the 'shop talk' that exists.

And how do you learn how to do successful welds?

Answer: TEST THEM!!!

Use one of the tubes in your images as a test-piece. Find the highest grade bolt in your workshop that will fit in the tophat-hole, and bolt the test-piece to a solid lump of steel, say a 1" thick steel-plate. Clamp steel-plate in the biggest vice in your workshop. Take a BIG hammer, or a heavy steel bar (say 1"+ round x 2'+ long), and bash away at the tube. You must keep going until either the bolt, or the weld, or the tube, parts company from the vice/big-steel-plate, and the end of the tube is lying on the floor.

If weld fails, then BAD WELD!

If anything else fails, then GOOD WELD!

Too easy!

Z

I actually did this, but on a smaller scale. I put the tube in the vise and beat on it a bit with a hammer. Since we didn't have any extra pieces made I didn't want to destroy the part for the car so I went a little easy. Nonetheless, they didnt fail. That experiment, coupled with the TTT diagram explanation gave me a bit more faith in the parts.

To my eye, you have several problems going on.

1) There should NEVER be slag or visible impurities on the surface of a properly performed tig weld, ever. Period.
2) Looks like insufficient shielding gas coverage - the color bands outside of the weld area should be farther from the bead. The bronze/gold color over several beads means they were either not properly covered (too little gas flow, too small of a gas nozzle, electrode protruding too far from the nozzle), possibly improper shielding gas. The bead should be very close to silver in color - some discoloration where you start or stop depending on pre- and post-flow of the shielding gas is ok.
3) Improper cleaning - that's where your flaky bead surface is coming from, it looks like the mill scale wasn't cleaned prior to welding. Especially with alloy steel which often has a light black oxide coating this is absolutely critical. You need to spend time with a metal wire brush or Scotchbrite to get down to clean shiny steel at least 1/2 in. away from your joint, then wipe with acetone or IPA just before welding - that alone will help a lot. Use only something that will not remove material - avoid sand paper, flap disc wheels, grinding wheels, etc.

ER70S-6 should be fine for that combination.

The welds are not perfect, that is for sure. It was likely a shielding gas issue. We replaced the bottle and have been welding re rest of the frame (4130 with ER70S-2 filler) with no issues (other than a healthy amount of heat input!). We are definitely paying more attention to the cleaning and preparation of the tubes. As Z said, ask 100 different welders a welding question and you will get 100 different answers. I know the welds are not perfect, but based on experience this type of low heat TiG welding on thin parts is strong enough for the application.

Thanks to all who replied and sorry that it took me so long to get to this. Between myself and the rest of the team we learned a good deal through this process and am glad that it occurred. Happy building!

[ryanjones2150]

I had attempted to reply to you way back when you first posted this, but the site was experience some difficulties and I was unsuccessful in my reply attempt.

I wasn't going to beat you up about the welds, especially since you say you're a student. Everything that has been mentioned is absolutely true. However, we can all look back on our welds early on and would surely see the exact same thing. You've been given some excellent advice on how to get better.

One thing I would like to point out, and I'm not sure if you mentioned it, is how important it is to clean the inside of the tube as well. By looking at your welds, there is an obvious "wrinkle" to many parts of the bead surface. That tells me you did not clean and prep your weld surfaces properly, to include the inside of the pipe. I use an end grinder with an abrasive wheel. It takes just an extra few seconds and can make a big difference in your welds. It is quite common to not get all of the mill scale off, which is most likely the case in your welds.

You seem like a highly technical type feller, and that's a good thing. If you pay as much attention to the practical application as you do the theorical application and I think you'll be ahead of most.