How a 3D Printer Can Identify Its Own Mistakes

October 27, 2015 | Author: Donald Godfrey

Honeywell is Partnering with Sigma Labs to Develop Objective Evidence of Compliance to Design Intent for Additive Manufacturing Technology

Normally our blog postings have been a short explanation consisting of a few hundred words describing what Honeywell is doing in the area of Additive Manufacturing. However, Honeywell is developing two system approaches on the topic of “Objective Evidence of Compliance to Design Intent” and will use two blog postings to describe this technology and the Honeywell effort.

As 3D printing technology moves closer to production application, the manufacturer must verify that the part meets the geometrical, metallurgical and mechanical property requirements identified by the design engineer.  Today in most industrial markets, inspection is completed after the part is produced and by mechanical technologies such as measuring via Coordinate Measuring Machines (CMM), X-Ray to inspect for internal defects, CT Scanning for looking deep beneath the surface of a part.  However, all of these technologies are subject to human error.  A person might not read the CMM results correctly.  X-Ray may only capture a possible void or crack near a surface.  CT Scanning technology is not widely available, and CT Scanning and interpretation takes considerable training to ensure proper results.

This is why Objective Evidence of Compliance to Design Intent is important.  Objective evidence is captured when no human is involved.  Therefore, an in-situ (in process) method of sensing and monitoring build parameters and part geometry is preferred.  Honeywell is working with Sigma Labs to develop two separate quality systems that will help make this possible.  Below describes how this technology is being developed and how it will work.

The way 3D printing works (in any machine) is the build software slices the part to be printed into thousands of layers.  Each layer (slice) correlates to the number of passes the 3D printing machine must make to build a part.

At a high level, a part to be printed is sliced into thousands of digital layers. For this example, we will print a cube and that cube will be divided into 3,000 equal slices (in the vertical direction) with each slice being the same thickness.  To print the cube, the 3D printer will make 3,000 passes to build the part from the bottom to the top moving in the vertical direction (referred to as the Z Direction).  With each pass or slice completed, the PrintRite3D® CONTOUR TMSystem will take a digital photograph of the slice just printed.  When the entire cube is printed the system will have taken 3,000 digital images.

With each image, the computer compares the image of the printed slice to the digital slice of the CAD model (file) at the same height in the part.  For example, if the system takes a digital image at slice 1,870 then it will compare that image to the CAD file at slice number 1,870. 

The above figure shows a digital image of one of the many slices taken when building a Honeywell engine mount.  The computer measures the features in this photograph and compares those measurements to the CAD file providing objective evidence of compliance to design intent.  If the computer identifies that the slice just printed deviates from the corresponding slice of the CAD file, the layer is reported as suspect.  If the difference in slice geometry is significant, the 3D printing machine may be programmed to stop printing that specific part in the build. The remaining parts continue to build.  If the difference in slice geometry is not significant, the 3D printing machine may continue building but send an alert message to the operator that the two digital slices might not be identical and should be considered suspect.  A suspect slice may or may not reflect a possible reject.  It is only an anomaly that should be reviewed after the build is complete.

The advantages of this type of quality system is that with enough empirical data a part can be produced and not be subjected to post build inspection techniques such as X-Ray, CT Scan or CMM measurements.  It is important to understand the concept of “enough empirical data” before conventional inspection methods are removed from the manufacturing process.

The objective of an In-Situ (in process) approach is not only to obtain objective evidence of compliance to design intent but also (in time and with enough data) to avoid the cost of inspecting the part after the build.  The above picture shows four Honeywell engine mounts and a digital fingerprint of the parts by slice.


The ability to generate objective evidence of compliance to design intent is possible with the correct technology and correct data gathering techniques.  This system is being developed via cooperation between two companies and will help Honeywell maintain its position as one of the global leaders in this area of research and application.  The next blog posting will reveal another approach to “Objective Evidence of Compliance to Design Intent”.

Donald Godfrey

Donald Godfrey

Donald Godfrey is an engineering fellow at Honeywell Aerospace.

Contact Information