Velo3D raised $22 Million in 2015 and was working in secret to revolutionize metal 3D printing. For the past years the company has been quiet as a mouse about its process and intentions. A lot of speculation abounded as to what Velo3D could unleash upon the world. Today we learn that the company has developed a metal printing process with more design freedom in metal. The company says that its systems can print complex geometries below 45 degrees. Which would make more 3D printed parts possible with their technology. The company has also developed its own software to acompany its process. And rather than just raising $22 million it turns out that they’ve raised over $90 million in funding. 3DPrint.com interviewed Stefan Zschiegner, Chief Product Officer at Velo3D, about the secretive start up now coming out of stealth mode. From their answers, their published work and patents we can conclude that this is a well captilalized start up with a lot of candle power that seems to have gotten quite far in controlling for many the important variables in metal 3D printing.
Why all the secrecy?
The model for Silicon Valley has typically been to announce and hype products long before they are commercially available. For a solution like the one we are bringing to market which aims to disrupt the $500 billion global manufacturing industry, we felt it was necessary to wait until we had a thoroughly vetted, customer tested product available for sale before announcing ourselves to the world.
Unpacking Acetabular cups (for giants) with the Velo3D.
What’s special about Velo3D?
We started Velo3D with a bold vision to enable additive manufacturing without design constraints. We are solving problems with deep insights and getting to the root cause. Based on that we build a solution from the ground up for high volume manufacturing consisting of our Sapphire System and Velo3D Flow print preparation software. Intelligent Fusion is the technology that powers the combination of Flow and Sapphire and enables an end-to-end integrated workflow.
While conventional systems often require supports for any geometry below 45 degrees, Velo3D’s Sapphire uniquely enables engineers to realize designs with overhangs lower than 5°and large inner diameters without supports.Some of our key benefites include
1st print success rates of 90%
Reduced part costs by 30-70%
Is this a manufacturing technology?
Yes, Velo3D is a metal additive manufacturing solution company. Our customers are service bureaus who offer metal 3D printing services to end users, as well as leading OEMs for use in-house.
What kind of parts can be made with your technology?
We have removed design constraints by enabling overhangs below 5 degrees and large internal openings up to 40mm. Key applications include shrouded impellers, heat exchangers, pump housings and other turbomachinery components which are critical for the aerospace, energy and industrial applications. We also enable medical instruments and implants, such as orthopedic hip cups.
You state that more geometries can be made? How?
The ability to design and print complex geometries is enabled by our Intelligent Fusion technology. Intelligent Fusion is a Velo3D proprietary technology invented to free the conventional powder bed laser fusion approach from design constraints through process simulation, prediction, and closed loop control.
Did you manage to correct for melt pool size in order to improve microstructure control?
Yes. Microstructure control is only one of Velo3D’s benefits. It allows us to build previously impossible designs and to improve part-to-part consistency.
What kind of reliability and repeatability are you getting?
We are meeting and exceeding reliability and repeatability tests by our customers. Currently we are testing with external labs and plan to publish the results soon.
How dense are parts?
The parts meet and exceed metal manufacturing density requirements of over 99.9%.
What kind of Ra are you getting off the machine?
The surface properties are geometry dependent and customer application defined. We are demonstrating below < 3 SA.
What post processing typically needs to be done?
The Velo3D solution minimizes the need for supports reducing typical support volume 3-5 times. It avoids internal supports that prevents the manufacturability or causes laborious post processing with conventional approaches.
Who are your target customers?
Service Bureaus and OEMs with expertise in additive manufacturing.
What are your target applications?
Aerospace, energy and Industrial applications, as well as medical applications (i.e., orthopedic implants). Applications include engine parts such as impellers, heat exchanges, and other critical turbomachinery parts, as well as assembly simplifications but also spare parts and spine implements, and larger implements such as hip cups.
Velo3D has come out of nowhere to seem quite the contender. If their estimates and performance claims pan out in the real world then this is a very interesting technology indeed. Simulation is very difficult to do in metal 3D printing and its a key element of getting prints right. Finding out $5000 and three days later that your parts don’t work kind of holds the technology back. This opens up new applications for 3D printing. Especially if they can have a success rate of 90% on the first time printed parts. Sometimes in Powder Bed Fusion you have to come up with different support strategies and print a part four or five times to get it right if it is a new geometry. Powder Bed Fusion in metals is great at making a million different hip cups but if we’d throw a radically different shape in the printer for the first time then this will most likely fail. Many applications are being held back because of this. Think of “draw your own jewelry” as a startup idea for example. Reducing supports will also make this much cheaper in terms of overall part costs and may save time as well. Supports are still manually removed on, nearly all, Metal Powder Bed Fusion 3D prints. You can see how parts are being unpacked on the Velo3D here. It seems that some post processing can be done on the machine but how much actual labor is saved is unclear. Manual removal of supports adds considerable cost to the final part so any gain here would be very beneficial. The increased design space could open up new applications, especially in new customers that have thus far been unable to make their parts with metal 3D printing.
3D printing is very much a testing and data game if you want to take on manufacturing. You’ll need hundreds of kilos of a powder to make sure it works well for example on one type of machine. There are also many geometries that can have significant effects on how and if the part builds. Thermal stresses can cause parts to get ripped apart as well. By developing simulation software the 100 strong Velo3D team has really focussed on getting the repeatability right through lowering their testing cost and increasing their dataset. This is a smart move and will bring dividends to them and their customers. The company has a number of patents including a skillful 3D printing one, an accurate 3D printing one and an adept 3D printing one. Also the first time I’ve ever seen cute patent names. Going by those patents the company has developed a real-time melt pool monitoring technology that works in concert with material dosing and laser control and builds closed loop using, probably, a plasma beam. The company seems to also to be able to correct on the fly with cooling to reduce deformation and may use an FPGA or similar to do this. So depending on errors, it seems to be able to reduce the intensity of the laser or actively cool a part. Given the teams previous work and published articles they may also be using a MOSFET or Field Effect Transistor to do this.
Also given that velocity fields play an important role in metal 3D printing and the name of the company is Velo 3D, I’m taking a guess here to say that they probably are managing to control velocity fields in some way which would then allow them to have more of a grip on the final part and how it is built. If they then are able to monitor melt pool size and shape in real time using the FPGA and then have an influence on the cooling rate of areas of the part while being able to adjust the plasma beam also on the fly, then they may have just come close to cracking this metal printing thing. They certainly have the candlepower to do it, they’ve hired very bright people who have over the years written some very interesting papers on real-time monitoring, modeling, and control of metal 3D printing. During my research for this, I was actually at one point surprised to learn that Brent Stucker didn’t work at Velo3D now given the overlap.
The patents also seem to disclose that parts can be polished and post-processed in part by lasers on the build machine, perhaps in concert with building the part. Another patent seems to point to active cooling or producing multiple layers at once using a preheat step or process. Can’t wait to find out how this actually works. All in all the value proposition seems a solid one and they’re certainly ticking the right goals. They’ve also got a lot of air which they can use to iron out the kinks in the chain. I like the fact that the company seems down to earth and isn’t all “startuppy” about everything. But, first impressions are first impressions and we work in an industry where an awful lot of machines and dreams have caught fire. We’ll have to wait to find out what the performance is as tested or experienced by the customer, it will cost me some beers at a trade show but I’ll find out for you guys.