Over in Reddit there was a photo of a car wheel that had five radial arms. They had all been broken and they had all been (shittily) welded. There was universal agreement this was unsafe (possibly illegal).
But there is nothing intrinsically bad about welds. Welding is used all the time to join metal pieces.
What type of load is a weld suited for, and what type is it not suited for?
I did some googling, so I’m totally an expert now.
From what I’ve read, aluminum welds are generally weaker than the parent material. In other words, that welded-up wheel, even with good quality welds, is likely to be weaker than it was before it broke, potentially by quite a bit.
With steels, a weld can be made as strong as the parent material, but this is not the case with aluminum. In most cases, a weld in an aluminum alloy is weaker than the alloy being welded.
“The weld isn’t as strong as the parent material, which a lot of people don’t realize,” says Frank G. Armao, director of aluminum technology at The Lincoln Electric Co. “The heat from the welding affects the properties of the parent material. Rarely can you make a weld that is as strong as the parent material when you weld aluminum.”
For the sake of argument, I made the table below based on info from ESAB (the world’s largest producer of welding equipment) that spells out exactly how much strength you can lose in aluminum from welding it.
In this table, I’m listing the minimum yield strengths. This is the minimum allowable point at which the metal will start to fatigue and permanently deform when you pull on it. Also, the type of weld being compared is a groove weld (when you weld it end to end from either side and properly bevel the edges before welding).
An Engineering Tips discussion on the strength of aluminum welds:
It sounds like if you plan ahead and use the right weld filler material and have the right parent material, there may be some post-weld heat treatments you can apply to improve weld strength, but the average shadetree mechanic is unlikely to be aware of these methods or be equipped to implement them.
It would be helpful for the discussion if you could link to the post or image in question. Asking “what type of load is a weld suited for” is an overly broad question because there isn’t really any way to answer that comprehensively without delving into the details of the structure. With regard to weld types and processes, there are several factors to consider with respect to the integrity of a weld joint in different load conditions, including (and most critically) fatigue loads which is where unexpected failures often occur even though the structure in question has survived initial proof loading.
Casual analysis often assumes that the weld is continuous and homogeneous with the parent material which has led to unpredicted failures because while a good quality, full penetration weld is often quite strong in the bead itself other factors to consider include the depth of weld penetration, weld filler material compatibility with the parent materials and resistance to corrosion, possibility for voids and inclusions in the weld, surface defects in the weld bead profile, the heat affected zone (HAZ) adjacent to the weld for cold worked and heat treated materials, and the impact of the weld geometry particularly with regard to fillet/bevel welds and particularly with one-sided fillets. The analysis of welds requires some detail understanding of welding process and material behavior that many structural analysts don’t know because it isn’t something that is really covered in engineering classes.
To that end, Omar Blodgett’s Design of Weldments is really the canonical bible on the design and analysis of carbon steel weldments, and while young engineers often look at the book with bemusement at the freehand figures and use of nomographs for calculations, it is inarguably the best presentation of methodology for evaluating weld designs. It does not go into welding stainless steel, aluminum, titanium, or other more exotic materials which require more detailed knowledge and treatises, and naturally doesn’t address the use of finite element analysis for evaluating the stress in welds, which also requires some special methods, particulalry when evaluating welds for fatigue life and profile variability. In general, it is wise to be conservative in evaluating strength and life capability of welds and especially pre-qualified welds that are not or cannot be non-destructively tested or inspected (NDT/NDI).
Stand by, coming right up. But I’m not asking so much about this one specific case as much as welding in general. It’s OK to weld components of a ship’s hull, but not a car wheel. I was thinking along the lines of good for compression loads, not good for shear stresses; good for static loads, not good for dynamic loads or high-vibration applications, or something like that.
I am not a welder nor a mechanical engineer, although I did take a materials engineering class in college.
Here is a hot link to the photo. I don’t know if you need a reddit account to view. If you can’t see it let me know and I’ll upload a copy to someplace.
It depends on the type of weld, the load application, and the structure in general. Welds almost never fail in compression and pretty rarely in direct shear (except for resistance spot welds, which shouldn’t be used for any critical structure); most weld failures occur due to bending or fatigue in cyclic loads; or because of incomplete penetration, inclusions, or voids affecting the nominal quasi-static load capability.
I can see the image and that looks like a cast (or maybe forged) aluminum wheel. There is no way any competent welder would have thought that was a good idea because not only does the wheel see complex flexural loads as it rotates with hundreds of millions of cycles in its lifetime but also high dynamic lateral loads. We can’t say how that failed originally but my guess would be a curb impact so even aside from the questionable welds but the underlying material probably has plastic deformation in addition to the damage to temper from the welding will probably make the interface to the welds highly brittle. Cast aluminum wheels are so cheap now it doesn’t seem like it would even be cost-effective to try to repair this damaged wheel.
While usually true, there is a type of welding for aluminum that can produce joints nearly the same strength as the base material: friction stir welding. It does this by mechanically swirling the material of two pieces together. Although the joint does heat up, it does not reach the melting point, and so avoids the problem of the parent material weakening. It’s used quite a bit in aerospace and is not a trivial technique.
Needless to say, whoever inflicted that monstrosity on the wheel did not use friction stir welding.
Friction stir welding only works on extruded aluminum in butt welds, and generally is only done on flat sheets or cylindrical sections because of process constraints of maintaining force application and dimensional control. While the strength of the material is comparable to the parent material form most tempers, you get a joint strength reduction of ~30%, primarily because it isn’t possible to get full penetration in the weld bead. It does produce very consistent welds with little variability in strength and geometry with no inclusions once the process is dialed in but it often takes a lot of effort to get the system set up for a particular joint, and issues such as tool wear, application force variation, et cetera can result in an unreliable joint so statistical process control and full inspection is necessary to avoid producing bad joints. Unlike normal TIG and MIG welding, if you get a bad joint in friction stir, it’s a complete scrap as you can’t repair the weld, so it can a very costly process.
Improvements in robotics are increasing the envelope of applicability for FSW. And it’s not just aluminum, though that’s probably the most common material due to the aforementioned strength loss with other methods. But yes, it’s mainly useful for butt welds, and is particularly useful for pressure vessels under tensile loads (as you mentioned, the strength loss from compressive or shear loads is rarely relevant).
Custom tooling for the stirring tool can also increase the scope of application, but that does make it less “generic” than, say, MIG/TIG welding. It’s a manufacturing technique, not a repair technique.
Looking at the wheel in question I think it would get you home and then you replace it. Portions of the welds are cold with little penetration. Looks like it was prepped with an abrasive wheel, another no-no as it introduces aluminum oxide into the weld. TIG welding require the use of stainless steel brushes for cleaning and the brushes are marked and not used on anything else. Aluminum cannot be easily heat treated to reduce stress like steel can.
Still, aluminum can produce beautiful welds under controlled conditions. I cracked the head on my 1971 BMW Bavaria right through the exhaust port. At the time a replacement head was about as much as the car was worth. BMW claimed the head could not be welded.
I went to our expert aluminum welder at work and we devised a plan. I carefully ground out the crack using steel burrs, no sandpaper. I wrapped the head in 2" alumina insulation and heated it slowly in a heat treating oven to very high temperature, I think around 500 degrees F. It welded beautifully and I used the filler rod the welder suggested. Back into the oven for several hours to slowly cool. I bought racing valve seats and had a regular auto machine shop install them. I put nearly 90,000 miles on the engine after that and had no problems at all…
Many 2000-series and 6000-series alloys can be heat-treated after welding just fine, and 6061 is not difficult to restore to T6 condition¹. It does tend to warp a fair bit, but the right jigs can substantially mitigate that effect. (The wheel in question is almost certainly cast, not forged; if so, it’s not 6061, so yeah, that weld is nuts).
I agree with everything you and Stranger said about the wheel. This bit about aluminum in general is the only thing that I didn’t follow. Can you elaborate on what you mean?
This has to have been started as a joke. Assuming that all five of the spokes had fractured so the hub had completely separated from the rim, as the photo suggests, there is no way such an assembly would survive highway speeds.
Any reduction in strength resulting from welding could be addressed by simply reducing the load capacity of the wheel. Much more difficult would be to ensure the rim remains concentric to the bolt (or stud) holes and perpendicular and true to the mounting flange. While the welds would definitely affect the material strength, I suspect the geometric issues would have a much greater effect on the suitability of such a repair than the material strength issues.
Something that continues to bug me about that picture. The tyre on the wheel is a Pirelli P Zero - so not exactly a cheap tyre, rather one for reasonably high performance cars. Looks fairly new too.
P Zeros are asymmetric tyres. They only work one way around. This tyre is clearly showing off the “Inside” label on the outside of the wheel.
Given the hilariously awful mess of the wheel perhaps this shouldn’t be a surprise.
I haven’t welded aluminum in so long I guess I am out of touch. My home TIG is DC only, for steel and stainless. The process for hardening aluminum used to require very cold temperatures for extended periods of time. Looks like they have it all figured out now.
A lot of corrosion resistant applications will use duplex stainless steels as opposed to austenitic. Welding duplex can destroy the corrosion resistance of the metal in the heat affected zone and can require the weld be annealed, so not a load problem in itself, but a durability issue for sure.
Many years ago I delivered some pipes to a field in rural Worcestershire. 10 pipes, each about 1.5m by 12 metres. They were soon lifted off and I went to the hut for some tea.
In there, with a group of rigger/welders was a small guy. Not a dwarf, but pretty short and i was told that his job was to inspect the welded pipe from the inside. I wondered how it was done, so they showed me.
The end of the last pipe was still above ground and when the new pipe was put in position, they were auto-welded together. Several pipes were added and this is where the little guy started work. He went down inside with some instruments and between them, they examined each weld for integrity, before they were all pushed into the trench and covered up.
But the vehicle used for drifting will likely wear the outside of the tyre. We can only see the original “inside” ?? SO they are using it backward to use the unworn part of the tread.
While it might be they could find welding rod material that is stronger than the alloy of the rim,
the trouble with these these aluminium alloys is that a small defect will develop into a crack.
If it was just going to be used as a hanger, say a few times a day, then it could last years.
The airplane people count number of take offs …
On the high performance vehicle, the rate of stress cycles is so large. those cracks develop quick…
Probably how it got this way first time… One bang on a gutter or something, and its got a little crack in a spoke, and then when that spoke is cracked completely through, the rest are overloaded and crack quickly.