Sigma Labs Inc (SASI)

[GetRich1day]

I went back to reread the whitepaper about the DARPA ICME rapid qualification framework. This is the probalistic framework that DARPA has IPQA built right into it and we are in Phase 3 now with completion mid 2018. I'm betting AM standards from FAA will be out right around the time this Phase 3 is concluded. IPQA being part of this framework that saves time and money is a great thing for SGLB. I see many orders coming for this PrintRite3D in this framework next year. Here's a few quotes.

Targeted testing and in-process sensing must be performed in order to inform the models such that they can be calibrated, validated, and verified. The models must be framed in a probabilistic setting in order to ascertain the sensitivity of the predictions to the parameters of interest and to understand the expected variability as a result of the
random nature of the inputs. At the same time, in situ process monitoring is being implemented in order to provide evidence in the form of In-Process Quality Assurance™ (IPQA® ) that the process is under control for the purpose of certification. It is this suite of activities and the knowledge generated that makes the rapid qualification possible.
The goal is to be able to reduce the qualification process time by 40 % and costs by 20 %. We intend to achieve the time savings by exercising the process and material models to predict the best processing conditions to minimize defects in the build and to predict the best heat treatment cycle to obtain the highest material strength. This will substantially reduce the exploratory set of experiments needed to define processing conditions and to optimize the heat treatment cycle. The cost savings will be the result of the substantial reduction in material testing needed to optimize the above processes.

A discussion of the different technologies that are needed in the rapid qualification process was presented, including the probabilistic design format, ICME models, and the quality assurance methods that form the framework. It was demonstrated how sensitivity analysis within a probabilistic format can be used to identify the important design variables as well as how model calibration can be used to improve the model accuracy. The process models were able to identify the processing conditions that can lead to defect-free manufacturing process parameter settings. The residual stress models were able to reasonably predict the residual stresses and deformation during the build process. The analytical models that were used to predict the alloy microstructure
as a result of post processing heat treatment were validated. These models were able to identify the type of experimental samples needed to characterize the material microstructure under processing conditions of interest.
The current status of the computational codes is encouraging, but there is a need to improve computational efficiency to enable simulation of the larger build spaces associated with real-world components. The tools and models are now available to simulate the DMLS® process for many material and processing conditions. They can be used to predict the process parameters that will yield good material consolidation with low porosity and to the top layer surface roughness from the predicted melt pool profiles. This leads to the capability to more narrowly defined process parameter conditions to perform a confirmatory series of specimens that is much smaller, thus saving time and resources especially during the metallographic analysis step, which typically takes a significant amount of time.
In-process monitoring is used to validate the models as well as In-Process Quality Assurance™. The sensors and corresponding IPQA® provide the link between specification and numerical process window selection to the actual build and final product. The uncertainty quantification tools provide a probabilistic foundation for decision making enabling the
assessment of overall process risk.
Lastly, while the models are described as fully integrated and somewhat sequential in the process, the use of them is independently valuable. For example, a process model can help define process limits for good integrity builds, independent of microstructure or strength. These models and this rapid qualification framework require, and will continue
to require, a substantial amount of “informed testing” to continuously support the analytical predictions and ensure qualification/certification authorities can have confidence in the results.

In order to enable rapid qualification, a holistic risk-based probabilistic framework is proposed. The models are utilized to identify the optimum process window. The applicable bandwidth of process controls and material property variations is estimated using uncertainty quantification. Manufacturing risks are mitigated by incorporating process monitoring and IPQA® ; the real-time assessment of build quality utilizes the range defined by models and uncertainty quantification to immediately inform the user of the manufacturing process status.
A high-level schematic of the probabilistic framework is shown in Fig. 2, where ICME is coupled with design models and methods. The inputs to the framework are requirements and critical-to-quality (CTQ) metrics, which are redefined as risk-based acceptance criteria to be consistent with a probabilistic framework. The preferred output of the framework is validation of design by analysis.

Another important benefit of a probabilistic framework is that once the validation testing is completed, the data gathered would be used to update/validate/calibrate the models. As a matter of fact, any additional down the line testing, such as certification and substantiation testing, would also be used to update/validate/calibrate the framework. This ensures continuous improvement of the frameworks’ overall robustness and predictability.
The framework presented in Fig. 2 contains within the dotted line box the set of ICME tools needed to achieve rapid qualification. A micro-model which describes the melting of the powder and solidification of the material to form the part, and a macro-model which describes the residual stresses that build up as a result of the rapid cooling and solidification of the material. Both of these models were developed by ESI to simulate the powderbed laser additive manufacturing process. QuesTek® developed the models to predict the material microstructure that develops as a result of the heat treatment process along with
a model to predict the tensile properties of the resulting microstructure. Sigma Labs and Stratonics developed techniques for in situ monitoring of the process. Sigma Labs uses a pyrometer to monitor the response of the melt pool whereas Stratonics uses a digital
imaging technique to capture the response of the melt pool. Furthermore, Sigma Labs has implemented their In-Process-Quality-Assurance (IPQA) technologies to determine if and when the process may be out of control. Honeywell is implementing non-destructive evaluation (NDE) methods to determine if defects that may develop in the part during the
manufacturing process can be detected by inspection. SwRI® is developing the probabilistic design and uncertainty quantification tools to be able to define the minimum predicted material property and risk derived from the novel manufacturing process. Lastly, Honeywell is developing the overall rapid qualification framework to bring all these technologies
together in order to demonstrate an acceptable means of compliance with
extensive use of simulation models.