AM Investment Strategies
AMS Spring 2023

Additive Assurance Mounts Quality Control Offensive Against Metal 3D Printing Defects


Share this Article

For years, metal additive manufacturing (AM) has been sitting on the bench, prepping to get into the production game. Most coaches just haven’t seen the technology as quite ready for the playoffs. There are exceptions, of course. Companies like GE have worked extensively to developed specific machines to produce specific parts, but it hasn’t been easy. Forgetting the basketball analogy, what’s needed is consistency and quality assurance to make AM ready to manufacture end parts.

As it stands, most metal 3D printers are not turn-key machines but require tons of trial and error to 3D print parts that don’t crack or become distorted and can do so repeatably. This results in a waste of material and machine time necessary to make up for the investment of a machine that may well cost over $1 million.

Tackling this problem has come from two directions: predictive software and quality control hardware. On the software side, companies like ANSYS have developed tools to predict and compensate for distortions in 3D printed parts. Using finite element analysis and other sophisticated methods, it’s possible to understand the physics of the metal powder bed fusion (PBF) and pre-distort components to make up for the changes that will occur in printing.

On the hardware side, new methods for examining and analyzing the build chamber as a print job is taking place are leading the way to better quality control in-situ. Up until relatively recently, machine manufacturers have relied on cameras and thermal sensors to measure the intensity of the melt pool and energy beam to provide some idea of what is occurring within the bed. Most attempts by OEMs have not been sufficient to ensure “first-time right” printing, leading businesses like Sigma Labs to develop its own hardware and software for the job.

While Sigma Labs has received significant funding in the aerospace sector via grants from America Makes and via partnerships with large OEMs, the company’s tools involve significant modifications to the printers themselves to incorporate the new quality assurance hardware.

This where Australia’s Additive Assurance comes in with its AMiRIS product. The startup, founded by Andrey Molotnikov, has developed its own hardware that requires no invasive work to the metal 3D printer itself. Everything is maintained in a compact unit that can be attached to the observation window of the AM machine. This is particularly important for industry certification purposes, Molotnikov told in an interview.

“Most of the companies require certification for many of their parts and the current certification procedures do not allow for any modification of the process itself,” Molotnikov said. By mounting the AMiRIS hardware onto the outside of the machine, it does not interfere with the printer itself, making it easier to certify parts and processes for specific industries.

Here the Additive Assurance hardware is mounted to a Sisma metal PBF machine. Image courtesy of Additive Assurance.

Like existing quality control hardware, AMiRIS does measure the intensity of the energy source within the printer, but it does so over time, the way long-exposure photographs track stars moving across the sky.

“When material melts you can capture the intensity, essentially the radiation that comes from it. We patented a very clever methodology based on long exposure,” Molotnikov told “When you dig into astronomy books, you can see that by taking pictures with telescopes over a certain amount of time, we can actually then decipher where the stars were, how they traveled, what speed, their size and so on. And we’re doing the same in laser metal manufacturing. So, the laser moves fast and we capture a long period of time using this long exposure. We can then extract more information than people trying to capture intensity as quickly as possible. This is sort of flipping everything upside down, but it actually works better.”

The software then uses machine learning algorithms to annotate defects within the layers and reconstructs a full 3D image of the part in a way similar to a micro-CT. It then alerts operators to any issues that are occurring during the build.

“Prints are made up of multiple thousands of layers. You don’t want the customers to go through each of the layers and look for deviations. Even if we’re indicating them, that’s still very painful,” Molotnikov said. “Instead, we’re using machine learning to annotate those defects within the layers, but then also reconstruct it to a full 3D image. Essentially like an x-ray over a hand to look at a broken bone, we do the same kind of thing with metal parts.”

Outside of his startup, Molotnikov is a professor at RMIT University in Melbourne, where much of the foundational research for the company began. Additionally, he sits on the boards of some of the standards organizations that are certifying AM parts for various applications, including aerospace and medical. In turn, he has insight into the issues currently being faced by the 3D printing industry with regards to quality control and certification.

A 3D image constructed of the build using Additive Assurance’s machine learning algorithms and software. Image courtesy of Additive Assurance.

Additive Assurance is currently working more closely with end users of metal 3D printers, in part due to the fact that OEMs often want to sign exclusivity deals that would prevent a product like AMiRIS to be available to competitors. However, Molotnikov suggests that his firm wants to aid the entire industry and will, thus, work with each customer as they come along.

The end goal for most companies working on quality assurance for 3D printing is to enable closed loop control, in which the system corrects itself during the printing process and, therefore, prevents errors from even occurring in the first place. Molotnikov would not say the extent to which Additive Assurance has gone down that road, but it seemed clear that it was something that the firm was exploring.

So far, the AMiRIS system has been used with a variety of machines, including 3D Systems, EOS, Sisma-Trumpf and GE Additive. The platform is currently available for purchase but won’t be delivered until next year as it undergoes some final cosmetic changes and testing.

Share this Article

Recent News

3D Printing News Briefs, October 1st, 2022: Flight-Ready Parts, Rapid Prototyping, & More

3D Printing News Unpeeled: 3D Printed MEMS, ASML and iCLIP


3D Design

3D Printed Art

3D Printed Food

3D Printed Guns

You May Also Like

3D Printing News Unpeeled: Boeing, Bauer, DBSchenker, Glidewell and Twikit

DB Schenker is to start a digital warehouse solution that lets you 3D print locally while they identify the right 3D printed parts, qualify them and qualify them. AM4AM is...

3D Printing News Briefs, September 29, 2022: Crowdfunding a 3D Printed House & More

We’re kicking things off with business in today’s 3D Printing News Briefs, as a Dallas construction startup is looking to raise $2 million to 3D print homes. LÖMI joined the...

3D Printing News Unpeeled: General Atomics, SLA on Textile, Dyze Design

General Atomics works with Conflux to 3D print  Fuel Oil Heat Exchanger (FOHE) for the MQ-9B. The company also says that it saves $2 million on tooling and $300,000 in...

3D Printing News Unpeeled: Formlabs, Hasbro, AddUp and Collins

Today we learn of Hasbro and Formlabs teaming up to make 3D printed selfie action figures. AddUp and Sogeclair will work with Collins Aerospace to make actuation components for aerospace....