Technology

Algoryx uses a systems engineering approach to solve molding problems.

Staying competitive is essential for survival. Maintaining a competitive advantage is essential for growth.	Engineers and managers on the cutting edge know that as competition intensifies, the need for greater productivity improvements, increased quality and substantial cost reductions continues to grow more critical. Even the very best of current techniques have not provided the quantum improvements required to outpace the competition. Until now
Profits, performance, and quality are degraded by design, tooling, process and quality engineers optimizing their respective subsystems instead of the overall system.	Historical gains in the quality of manufactured parts over the last two decades have been significantly eroded by increases in measurement and recording costs and by Statistical Process Control (SPC) and Process Capability (Cpk) analysis costs. The efforts of engineers to modify pre-production tooling and fixtures to production tooling and fixtures are frustrated by operator changes to process settings. Process engineers have difficulty determining the values of process settings and design engineers find it difficult to design for producibility when there are multiple part characteristics. The use of standard design tolerances increases manufacturing costs. Determining tolerance relaxations is a trial-and-error process. Replacing obsolete materials can be problematic. The inability to produce parts at design target values decreases product performance.
A Systems View of the Problems
Process complexities can make it difficult or impossible to understand the relationships between process causes and effects.	Process Complexities Process engineers and operators find it difficult to determine the process settings needed to produce a good (meets all specification requirements) first article when the following process complexities are present: More than a few process variables; More than a few critical dimensions (part features or performance characteristics); Differing dimensional responses to changes in any process variable; Processes that produce multiple units (multiple mold cavities or lay-ups); Simple or complex interactions; and, Non-linearities and reversals in dimensional responses to changes in process variables.
DOE doesn’t do the job in many cases.	Design of Experiments (DOE) has proven to be a useful tool for circumstances where (i.) one part or performance characteristic needs to be optimized and (ii.) there are few process complexities. When the first criterion is not met, DOE generally gives conflicting results. When the second criterion is not met, it is difficult to model the process and get useful results.
It is valuable to know when it is difficult or impossible to produce good parts.	Small or Non-existent Producibility Windows The producibility window may be non-existent. I.e., it can be impossible to produce good parts. Operators and process engineers can waste significant time learning this. Even if a set of process settings can be found that produces good parts, the producibility window may be so small that even minor variations in process settings result in bad (don’t meet specification) parts.
Design targets, design tolerances, tooling and process settings are interrelated.	Multiple Design Targets Cause Problems For products with multiple dimensions, process engineers and operators have the difficult task of attempting to adjust process settings to produce parts with all dimensions simultaneously at target values. This usually requires tooling, fixture or design target changes.
Operator changes to process settings can invalidate tooling modifications.	Problems Modifying Tooling Tool and fixture modifications can be risky and time-consuming. For all processes, dimensional results depend on the values selected for the process settings. This creates a problem for the engineer when deciding how to modify the tooling or fixture. Use of different process settings can invalidate previously made tool or fixture changes. Tooling and fixture engineers refer to this as the “tyranny of the operator.” There has been no prior art technology that tells the tooling and fixture engineer how to adjust tooling and fixture dimensions to transition from a pre-production to a production tool.
Operator changes to process settings can invalidate tolerance relaxations.	Problems Relaxing Tolerances The amount that design tolerances must be relaxed depends on the values selected by the operator for the process settings. Different process settings require different tolerance relaxations. There has been no prior art technology that (i.) solves this problem, or (ii.) gives the design engineer a ranking of the order in which to relax design tolerances or (iii.) the size of the required tolerance relaxations independent of process settings.Changing Process Technology When tooling or fixtures can not be adequately modified and/or tolerances can not be adequately relaxed for a given material, then a more capable process must be used to produce the part. There has been no prior art technology that enables the design, tooling/fixture, process and quality engineers to easily determine when this is the case.
Current SPC techniques are redundant and inefficient.	Current Statistical Process Control Studies are Inefficient During part development (qualification and certification), SPC studies are usually done on all dimensions. During production, SPC studies are usually done on either all or a subset of dimensions. As will be shown later, this is inefficient and incurs unnecessary cost.
Current process capability analysis techniques are redundant and inefficient.	Current Process Capability Studies are Inefficient In a similar fashion, process capability studies are done on all dimensions. This is also inefficient and incurs unnecessary cost
Current inspection techniques are redundant and inefficient.	Shipping and Receiving Inspections are Inefficient In a similar fashion, shipping and receiving inspections are usually performed by sampling a subset of parts and measuring all critical dimensions. This incurs unnecessary costs for both the customer and supplier.
A systems approach is needed to integrate design and tooling for manufacturability.	Problem Summary Design, process, tooling and quality parameters are interrelated because they co-jointly influence part characteristics and consequently how the part performs. Prior art has not provided engineers and decision-makers with an integrated system of technology that incorporates these interrelationships.
The New Technology
No changes are required to the manufacturing process.	In 2000, a large, international OEM was having difficulty producing good first articles for a new product line of injection molded plastic parts. These problems stimulated the development by Algoryx of innovative technology that solved the problems. The new technology has been proven in several industries with numerous case studies. No changes are required to the manufacturing process. The algorithms are implemented in computational software called Correlation Master. The software is easy to use.
Measurement and analysis costs have been reduced by an average of 90%.	Huge reductions in measurement and analysis costs have been achieved, as well as increased quality, increased productivity and reduced time-to-market. The bottom line has been significantly increased profits.
The new technology is implemented through powerful computational software.New process laws dramatically increase our understanding of manufacturing processes.	New Algorithms and Process Laws The pioneering technology consists of a system of interrelated algorithms which integrates techniques from the fields of correlation, regression, prediction, constraint theory, and statistics. It has led to breakthrough insights into manufacturing processes and to powerful new computational software. The pioneering technology consists of a system of interrelated algorithms which integrates techniques from the fields of correlation, regression, prediction, constraint theory, and statistics. It has led to breakthrough insights into manufacturing processes and to powerful new computational software. The insights are summarized in an original set of process laws (available on request).
Process complexities can be eliminated by ignoring the relationships between causes and effects!!!	Foundation of the New Algorithms The new algorithms are based on the fact that although the relationships between causes (process settings) and effects (part characteristics) may be difficult or impossible to determine, the relationships between effects for many processes are consistent and predictable irrespective of changes in the process settings.
	Figure 1 illustrates a condition where there are four correlated dimensions on a part-A, B, C and P-where P is the predictor dimension. Assuming perfect correlation, the relationships shown in Figure 1 represent the entire universe of possible part relationships. The relationships are defined by regression lines fitted through manufacturing data.
The predictor is the statistically best predictor of all other dimensions.	The predictor dimension is the statistically best predictor of all other dimensions. Determining the best predictor is computationally intensive.
The output from many manufacturing processes has a single degree of freedom.	The system of relationships in Figure 1 has a single degree of freedom. When the value of the predictor dimension (P) is known in Figure 2, the value for all predicted dimensions (A, B and C) is known.
	In the real manufacturing world, perfect correlation seldom occurs. Measurement error is one of the main contributors to imperfect correlation. The condition of imperfect correlation is resolved by bounding the regression lines in Figures 1 and 2 with upper and lower (three sigma) prediction intervals as shown in Figure 3.
	Figure 3 also shows the upper and lower specification limits for a predicted (critical) and predictor dimensions. The intersection of the target values for the dimensions is defined as the target intersection. The area within the four specification limits is defined as the region of conformance. It is not possible to produce a part with both dimensions at target unless the target intersection lies within the prediction intervals.
	Tooling Modifications The regression line can be shifted so that it passes through the target intersection. This is done by modifying tooling or other pre-process dimensions. The offset is defined in Figure 4 as the vertical (or other) distance between the regression line and the target intersection.
Adjusting tooling dimensions makes it possible to produce parts at design targets.The operating point is adjusted by changing process settings.	Operating Point Adjustments The operating point is defined as a point on any of the regression lines. Returning to Figure 2, when the operating point for any one dimension is determined, the operating points for all other dimensions is determined. The operating point is adjusting by changing one or more process settings. This is equivalent to sliding a bead along a wire.
	Four Possible Relationship Conditions There are four possible relationship conditions. The first condition is defined as “robust” and is illustrated in Figure 5. The critical dimension will be within specification limits regardless of whether or not the predictor dimension is within specification limits. The robust critical dimension never needs to be measured.The second condition is defined as “non-constraining” and is illustrated in Figure 3. When the predictor dimension is within specification limits, then both the critical and predicted dimensions will be within specification limits and the non-constraining critical dimension never needs to be measured.
	The third condition is defined as “constraining” and is illustrated in Figure 6. When the predictor dimension is between Pmin and Pmax, then the part will be a good part and the constraining critical dimension never needs to be measured.
	The fourth condition is defined as “defects” and is illustrated in Figure 7. When at least one prediction interval lies outside the region of conformance, defective parts are likely to be produced and the defect critical dimension never needs to be measured. Instead, tooling and/or design tolerances need to be adjusted.
	Operating Range Figure 8 is a one-dimensional representation of a part that has three dimensions-two critical dimensions (C1 and C2) and the predictor dimension (P). The top arrow represents the range of P for which C1 will be in specification. This information is obtained from a figure such as Figure 6. The middle arrow represents the range of P for which C2 will be in specification. The operating range is the range of P for which all three dimensions are within specification. This logic can be extended to any number of dimensions.
The operating target is the point that maximizes producibility.	Operating Target For symmetrical process output, the operating target is located at the center of the operating range. For non-symmetrical process output, the operating target is best selected as the point at which there is equal area in each of tails of the process distribution outside of the operating range.
Existing DOE data is a treasure-trove of untapped information.	Existing DOE Data Existing DOE data can be further analyzed with the computational software to obtain the many benefits outlined in this white paper. In this context, the new technology is “one large step beyond DOE.”
Following these steps leads to lower cost and shorter development times.	NPD Team Steps The New Product Development (NPD) team follows these steps to achieve the desired Cpk values during Qualification, Certification and Validation: Make and measure the induced variation parts Run Correlation Master software in the analysis mode – Determine the Operating Range – Determine the “as is” Cpk/Opk value using stable press output – Identify options to achieve the desired Cpk/Opk Have NPD team meeting – Reach consensus on which options to implement Run the Correlation Master software in the simulation mode – Confirm that team choices give the desired Cpk Make confirmed changes – To targets, tolerances, tooling, process, measurement Do Qualification/Certification/Validation/Production – Make parts that meet Cpk requirements
Increase Profit, Quality and Competitiveness
Measurement and analysis costs are greatly reduced during production.Producibility and quality are maximized during production.	During Production Greatly reduce measurement costs; Potential to eliminate destructive, reverse imaging, difficult and costly measurements; Greatly reduce the number of SPC and Cpk analyses; Simultaneously optimize system-wide Cpk’s; Operate at the point of maximum producibility and quality; and, Enable the operator and process engineer to concentrate on only one or a few dimensions.
Good first articles are obtained faster, cheaper and with less risk.	During Development Increase speed-to-market and decrease costs by not wasting time or money on impossible or extremely difficult to produce parts; Determine the trade-off between part performance and producibility; Instantly determine when to refer problem dimensions to the design engineer for tolerance relaxation; Get a ranking of design tolerances from most constraining to least constraining; Get the matrix of design tolerance relaxations required to achieve any increase in operating range; Instantly determine when to refer problem dimensions to the tooling engineer for tooling modification; Get the required changes to tooling to produce parts at design target values; Isolate operator process setting changes from the calculation of the required changes to tooling; Increase ability to produce good first parts by focusing on only one or a few dimensions; Increase the process engineer’s ability to determine the “sweet spot” for process settings by using the operating target and eliminating the effects of process complexities; Instantly determine the optimum material selection from an array of choices; and, Use a systems engineering approach that integrates the activities of the design engineer, process engineer, tooling engineer, manufacturing engineer and quality engineer.
Correlation Master simulates changes to the design and tooling without the time, cost and risk of actually doing so.	Simulation BenefitsThe Correlation Master software also simulates the operation of the manufacturing system used to produce the parts and the measurement system used to measure them. The Algoryx technology enables assessing the impact of contemplated changes to design targets, design tolerances and tooling on manufactured part dimensions without incurring the cost, time and risk of changing tooling or design parameters and then producing, measuring and analyzing the new parts.
Summary
Correlation Master is easy to install, uses standard operating systems and is user-friendly.	The Algoryx technology is a new model for manufacturing. It provides an immediate increase in profitability and market competitiveness by reducing cost and increasing quality during production and by reducing time, cost and risk during development.
	Algoryx, Inc.™ is the exclusive U.S. distributor of the Correlation Master™ software.Contact Information Algoryx, Inc. Los Angeles, CA 90049 310-820-0987 310-820-1046 Fax www.algoryx.com info@algoryx.com

Take the Algoryx Challenge – See the results using your production data!