A use for Large Metal Additive Mfr with Weld Wire? 1

[JNieman]

I've been researching more and more about Wire Arc Additive Manufacturing technology. The state of things seems to be pretty respectable, but not widely adopted as far as I can tell. The technical aspects all make sense. The final material quality, with good weld procedures, wire, and environmental conditions all make me think this is a no-brainer.

So what I'm wondering is... for people who end up working with / finishing low volume castings and forgings... are we going to start seeing "WAAM" near-net shapes to finish-machine rather than forgings and castings? Maybe see that as a standard "blank/stock" production method for the expensive exotic metals?

I know of a lot of production in my history that involved starting with a large cube and putting 95% of the mass in the swarf dumpsters.

The lead time on some of these low volume castings we end up sourcing for our customers can be atrocious. If we can "print" a part per day, even, it seems like turnaround time would be pretty greatly improved.

Has anyone else been aware of this type of thing?

Tangentially, I was at Fabtech last year, and Lincoln and Fanuc were certainly showing it off.

I'm curious how much sense it makes to bring in-house, or even to outsource to Lincoln, since they offer it as a service. Anyone out there have experience with it, or considering it?

 

[EdStainless]

It can be done, look at Sciaky.
The issue with this method and all other AM is that today no one can tell you what your properties will be (in various directions) before you build a part (and destructive test it) and where in the part you defects are (measured during fabrication).
Once you overcome both of these I'll be a lot more interested.
Today most AM 'success' stories are related to moving a conventional part that was over built by 50x and making it in additive at 25x overbuilt. You cut the weight in half, but still way over sized so defects are not an issue.
Or it is a new part that has special unique geometry and you live with it being over built.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[JNieman]

Sciaky is a great example. They transitioned from welding technology to additive successfully, it seems.

I had the reservations you did, as well, when I first started looking into additive processes to merge into current workflows and design possibilities. It appears to be that process control is key, like anything else. I think with the right information (even if it requires real-time recording of parameters during the buildup) of parts, quality can certainly be maintained and proven. I know I've sat through seminars on welding technology, where a process was developed, tested, proven, and was able to shed the inspection requirements. So long as power supply parameters were within a certain threshold, Ford considered it a good weld. I imagine a similar thing could be done for an additive process. Like any manufacturing plan, if you create a part, it gets tested and approved, and you make subsequent parts to the same specification, with the associated inspection plan, that's a good part. I guess my thinking is that, if we can do all this for arc welding, why not for arc-deposition?

Like weldments, maybe until we know more, basic NDT methods still apply.

What you say about the things being overbuilt (or worse, the psuedo phrase 'over-engineered') rings true. Just because it didn't break, doesn't mean it's well designed.

I think Sciaky is a good example of the potential, though. I don't see a reason it couldn't be scaled down (economically) to use the many ubiquitous metal wire stock available with a more typical arc-welding method though. That would put it in reach of a lot more applications, where an electron-beam system might be out of reach or too niche.

Another side of me like the idea of bimetal stock pieces, too, especially in tooling. Use the expensive exotic stuff only where you need it. Instead of large molds made 100% out of H13, do more of an "H13 clad" part out built up mainly from cheaper steel where possible. I could take a team out for a lot of lunch drinks with the leftovers in the budget from that savings alone.

 

[dgallup]

I've always outsourced rapid prototyping because my needs vary tremendously project to project and the tech advances so fast that it's never made sense to invest capital in it for anything beyond little desktop 3D printers. I'm sure plenty of others can build a great business case to bring it in house.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[JNieman]

Too true, re: pace of advancement.

One big difference between traditional machining and additive machines... you can throw just about anything in a CNC mill and it'll cut it up. AM machines are much more restricted to materials, I believe. I don't know if that can change much. That was part of my interest in wire stock like Sciaky and a couple others use. Man has been welding with wire feeders on just about any metal for a long time.

Looking at some of the research findings, the metallurgical properties of WAAM made parts is pretty respectable.

It's too bad you can't just rent or lease a machine like that to try out, just to see if it's a fit. What's the software model for that? Software as a Service? Is Machinery as a Service a thing? That'd sure bring down the barrier to entry and cost of trying.

 

[EdStainless]

Am is growing up. People are starting to work on special alloys for AM, similar to the way that cast versions of alloys are different from the wrought ones. The post processing is also a big deal, the HIP and HT is not cheap.
But properties are still an issue. Especially when we talk about fatigue, impact toughness, corrosion resistance, and any directional nature of properties.
The defect issue is a real concern of mine. If you are building parts with less over-design then defect type, location, and size become more critical. In order to get there they need to do a couple of things, one is to run closed loop with the process aware of fluctuations in speed, feed, temperature and such. People don't do this because if you are trying to control heating and cooling rates at all locations in the build because the build rates go way down. The other aspect is to know the locations of potential defect. I have worked with some AM Ti parts where the cost of doing CT on the parts was 5x the cost of building them.

There are firms out there will build parts as a service, but it is a new approach.
Using AM for tooling is huge potential, especially your bimetallic comment. The ability to rough machine a low alloy steel blank and then coat it with the correct amount of tool steel offers great potential. But if you are using H13 for hot work then selecting the base alloy that will continue to provide the needed support at working temp becomes a big deal.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[JNieman]

Yea, the places doing "job shop" type printing of parts is pretty well set when you're talking about plastics, nylons, and small metal parts. We currently have some of our standard designs outsourced through Xometry for SLS production from nylon powder.
When we start talking about large parts, possibly dinner-table sized - scale of such a service is not so simple.

I'd have to crunch more numbers to get even a back-of-a-napkin estimate on whether bimetal tooling is truly a major improvement in efficiency and cost, or if that's me thinking it's just such a cool idea. Sometimes such things are both. It's very interesting though, and it seems, like you mention, there's lots of potential, if well planned and executed.

AM in general though, I think is not just growing up, but mature. We're at the stage where improvements are starting to become cost-reduction to make it more accessible to general manufacturing, rather than the GE, Boeing, Lockheed, Bugatti, SpaceX, and Volkswagons of the world.

Thanks for the input all. I welcome any other thoughts, and I'll certainly report back if this goes anywhere for me.

 

[EdStainless]

Mature, not by a long stretch. But is past peak BS and reality is taking hold.
The day that I can pass a print and property requirement (UTS, yield, elong, RA, impact, fatigue, corrosion, creep or stress rupture) off and then actually get that part it will have arrived.
I seriously doubt if AM (from wire or powder) will ever be able to handle jobs that we use re-melted (high cleanliness) metal for today.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[JNieman]

Well, see, that's just the thing that prompts me to lean toward its state of maturity rather than simply being in development. You can certainly call out any spec you require for any other material stock and require a part made via some AM process comply.

I was just reading about ExOne, a company that does a binder-jet-process for AM production, meeting a material certification used for typical MIM production. In this case, they stated a third-party testing that resulted in verification that the supplied part met ASTM B637-18.

While that's a much different process than what I'm looking at specifically, it is certainly a decent example of maturity in an AM process material properties.

There are specific ASME/ASTM teams developing standards for testing where gaps may exist, but like most things, industry leaders are paving the way, what with Lockheed, Boeing, BAE Systems, etc all providing internal standards for materials and processes that meeting company spec. Meanwhile, there are existing materials specs that are certainly relevant and do not need rewriting to apply to additively manufactured products.

 

[EdStainless]

Raw material specs and testing are fine, but as long as we are in a world where I can't change material supplier, or machine, or even build orientation on the same machine and get the same results then you do not have a real process yet.
These guys wish that they were as repeatable as investment casting.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[dhengr]

JNieman:
In today’s manufacturing and fabrication, we are already doing/using AM. We just sometimes use rolled sections, coiled steel, plates, forgings and castings, etc. to save hundreds of miles of welding wire, of a material spec. which we can’t purchase.

 

[JNieman]

@dhengr
I think it's safe to say that is not the same nature of AM that most people are discussing when we discuss additive manufacturing

 

[EdStainless]

There is an area of activity that I have taken interest in, that is taking near-net-shape parts (forge, cold drawn shape, extrusion, investment cast, MIM) and using AM (powder fusion, NC weld, friction weld, or whatever) to build up special geometry details that you can't do with the original process. This limits using the more expensive/slower process to where it adds the most value.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[arunmrao]

Metal3D printing is rapidly progressing. Some of them are pure marketing hype. HP have built a resin-coated powder machine. I had seen it about 2 years ago at Univ of Kentucky Louisville. They also have other metal fusion printing equipment.