Injection molding process validation and design of experiments (doe)

Excela Mold™Rapid Injection Molding Commercialization and

Optimization

By Matthew Manges

What is Excela Mold™?

Excela Mold™ By Matthew Manges

Excela Mold™ is a proprietary mathematical modeling system that reduces mold Production Part Approval/Process Validation time and resources on average by one fifth. Many times much more.

Not only does it reduce process development time by one fifth, it automatically optimizes all variable process parameter interactions for best mathematical part dimensions down to .001 inch with predictive modeling.

How does Excela Mold™ improve upon RJG Scientific or Decoupled Molding?


RJG Scientific Molding method targets fluid and thermal dynamics, controlling pressure and flow thus maintaining shot consistency.

The RJG Scientific Molding method concentrates on making the same part each time (reduced shot to shot variation) and reduced cycle time but not whether the part is in specification or the process is dimensionally capable.

Excela Mold™ Advantage Summary


Custom System Procedure and Work Instructions

dependent on type and scope of Quality System designed specifically for the Injection Molding industry.

Automated scalable Flow Chart decision or trigger points for poor capability and performance. Depending on modeling results, could point to Metrology or other target specific areas.

Custom variable interaction modeling trade-offs. Has the ability to reduce other Cpk’s to increase a problematic one to >1.33.

Cycle time reduction. Getting it right the first time.

Excela Mold™ Case Studies


The best way to help understand the Excela Mold™ process is to walk through several case studies of the following;

Body Wash Cap Deep-Cycle AGM Battery Case Deep-Cycle Flooded Battery Case

Body Wash Cap Case Study


Body Wash Cap Case Study


This was a new product with a refurbished pre-existing cap prototype mold fitted to a new bottle design. This cap with new dimensional specifications, proved to be a major challenge for the Process Engineer and was unable to fulfill product requirements.

The following is a raw data output of the first step in the Excela Mold™ system, a screening DoE (Design of Experiments).

Understanding the Excela Mold™ Screening DoE Modeling Basics


The column on the left are the target dimensions. The “y” output is the best possible dimensional solution for all the given specifications at the same time. The “d” statistic is very important and tells you how much of the tolerance you can use for a given dimension while keeping all other dimensions optimally aligned (0 to1, 1 being perfect).

The top column has the run condition ranges for Hold Pressure, Hold Time and Cure Time. The numbers in red are the Optimal machine settings for the given specifications.

Body Wash Cap Case Study Screening DoE






Immediate benefits of Excela Mold™ process screening DoE Modeling


Simple complexity management. There is no way a Process Engineer could manage this process with 30 different positive and negative slope interactions without modeling. Let alone dial it in with .001 inch precision. Basically, without initial DoE modeling, the Process Engineer is guessing or “feeling around in the dark.”

The Initial Excela Mold™ process screening DoE will feed into other modeling and critical decision points within the Excela Mold™ Part Approval Process or IQ/OQ/PQ (which will be discussed later in the presentation).

Excela Mold™ Body Wash Cap Case Study Summary


Day 1- RJG Scientific Molding cavity balancing (4 cavity), rheology curve, gate freeze and fill rate optimization completed. 16 part screening DoE completed.

Day 2- Metrology and DoE modeling completed. Optimized process verification capability sampled (30 samples).

Day 3- Optimized process verification capability run passed all PPAP and CpK requirements.

Production – No non-conformances were ever generated with prototype tool while production was being built.

Deep-Cycle AGM Battery Case Study


Deep-Cycle AGM Battery Case Study


This was a new battery product design with a new production mold. AGM batteries in general require specific within cell plate compression for operation. The bottom cell dimensions were critical and tolerances were changed and tightened after 1st DoE screening.

The following is a raw data output of Excela Mold™ process target specific DoE (Design of Experiments).

Deep-Cycle AGM Battery Target Specific DoE


Deep-Cycle AGM Battery Target Specific DoE


Excela Mold™ Deep-Cycle AGM Battery Case Study Summary


Model Analysis- Although analysis showed we were able to keep the 2 outer cells within the new specification we would not be able to for the Partition feature target .010 USL .020 with y=.0189 and d=.11053. Further Excela Mold™ modeling showed that even with good Metrology Gage R&R and shot to shot repeatability, steel change was necessary due to predicted Cpk <1.33. Mold was re-sampled to confirm y=prediction (3 samples).

Steel Change – Model confirmed correct. Steel adjusted to .010 nominal based on the n=3 conformation samples.

Deep-Cycle Flooded Battery Case Study




This was a existing set of battery molds. The customer introduced COS automated manufacturing lines at it’s facilities. Battery end bow or warping had to be significantly reduced for the COS and a new specification was implemented. This was a very large account and with the new specification, had to rent another warehouse for heat tunnel/fixture rework. They were unable to meet the specification (%100 non-conformance rate) for months and filled this warehouse with non-conforming product.

The following is a raw data output of Excela Mold™ process target specific DoE (Design of Experiments).



Deep-Cycle Flooded Battery Case Study Summary


Model Analysis- The screening DoE showed each cavity part end walls (2) had a bimodal distribution regardless of Hold Pressure, Hold Time and Cure Time adjustments. The spread was so great that one wall would be at LSL and another at the USL. It was observed in the DoE, water (cure time) could change end walls. Separate water circuits on two end-walls showed promising results (previous slide).

Production– Separate die heaters were implemented for each end-wall for nominal adjustment. Rework went from %100 to %5 for the 6 volt and 8 volt cases.

Excela Mold™ Systematic Approach


The Excela Mold™ System was developed over time with the experience of 100 plus real world DoE’s and scalable escalation process actions commensurate to target objectives (mathematical decision points). Most of the techniques such as DoE are new to the molding industry and not understood as a whole.

For example, two major Injection Molding educational institutions recommend using fill rate as factor and Taguchi orthogonal arrays for Injection Molding DoE. The Excela Mold™ System shows that this is the last thing anybody should do, if ever in injection molding for the following reasons.

Excela Mold™ Systematic Approach


DoE and Process Modeling involve high level inferential statistics or mathematics. One needs to have a high level of understating of which type inferential test to apply to specific data characteristics. Injection molding is not like making semiconductors. There are 3 primary drivers with high signal to noise ratio in regards to Injection Molding DoE.

This high level of signal to noise ratio requires 1/4th the DoE samples you normally see in the Injection Molding industry. In regards to fill rate as a factor, on the left side of the rheology curve you have shot to shot inconsistency, far right, polymer shear and splay. Generally the fill rate DoE confirms the screening DoE, change steel or specification.

Excela Mold™ Contact and Pricing


We would need to do an assessment and discuss the right options for your company;

ISO 9001, 13485 or 16949? Emergency Validation Direct Hands on Mold Validation and Training Custom Procedures Practicing RJG Moldmaster I and II techniques? Do you have MiniTab or JMP software?

Please call 336-860-8005 or email: [email protected] for assessment. www.excelamold.com

mailto:[email protected]

http://www.excelamold.com/

About the Owner


Matthew Manges- After college, I spent 6 years in the Air Force. My first job after military service was in the SPC/Product Development department in the semiconductor industry. I initially worked under two PhD Statisticians and then worked on joint special projects with Micron and Intel. Worked with various medical device and injection molding companies for the remainder of my career. Consulted full time for 5 years for Zimmer, Boston Scientific and Stryker Orthopedics. 20 years total Engineering experience.

Quality Certifications ASQ-SSBB, CBA and CSQE