Improved engineering plastics with better material properties have led to an increase in the use of plastic molded gears in mechanical drives. The cost efficiency of replicating geometries through molding has also increased their use. With this growth has come greater demand for delivering new drives to the market faster. Reducing the number of steps in mold development reduces product-development schedules; one way to reduce steps is with careful, systematic design work that hones in on the best solution quickly. For precision plastic parts (and in particular, plastic molded gears) analysis with software significantly accelerates the development of fully qualified molds. Here we'll explore one such analysis system called the genetic molding solution, which eliminates the art component of final cavity prediction replacing it with a scientific approach to cavity definition through direct calculation.

The process of defining a gear tool cavity begins with the desired profile, which is typically scaled up by a linear mold shrinkage coefficient provided by the material supplier. Though common, simply scaling up a mold's size is not appropriate for making precision gears because shrinkage rates for the outer diameter, root diameter, and tooth thickness all vary. (Refer to pending AGMA Standard 909-AXX, Appendix D for more information on this.) The actual shrinkage depends on the particular gear geometry, material properties, and molding process used, and is defined experimentally. These empirically derived shrink rates can be applied to the critical gear dimensions to create the desired part profile. Called the differential shrinkage method, this technique works well if the gear body is simple, such as a flat, uniform disk. (In this case, the injection gates should be symmetrically located around the center of the gear.) But not all gear parts are simple in shape; in fact, many are quite complex. This is true for a couple reasons. To increase product reliability and take full advantage of plastic molding's cost efficiency, designers often combine gears with other components into one integrated part or one over-molded assembly. These complex parts often require that injection gates be located outside the gear center. In addition, large gear bodies often require alterations — ribs, spokes, or cored holes — to reduce plastic mass and avoid surface sinking, voids, and cracks. These additional design requirements limit the success of the differential shrinkage method.

The definition of the gear mold cavity profile becomes a serious challenge that is often addressed through trial and error. With this “educated guess” method, each successive mold cavity is designed to compensate for errors introduced by the previous one. This can take a considerable amount of time (not to mention several mold cavities) to achieve the desired gear profile. To speed up mold qualification, trial and error is sometimes combined with tweaking of the molding process parameters. Success of this approach is greatly dependent on the molder's experience with similar parts from similar materials. This means every new gear configuration and new material presents a new challenge — so with this technique, mold cavity definition is more of an art than a science. The need to make several iterations of the cavity during the molding process development wastes time and resources.

There are more scientific methods to define the final cavity profile. They typically use finite element analysis to model plastic material flow, mold configuration, and the molding process. However, these approaches are marginally successful in defining the final cavity of precision parts; accurately defining some of the factors is difficult, particularly the properties of molten plastic and the dynamics of the molding process.

Systematic approach

Similar to how DNA contains genetic information about the entire organism, the shape of the molded part reflects information about the original designed profile, material properties of the plastic, tool design, and molding process parameters. In the genetic molding solution method, all of this information is used to define the final mold cavity. Mathematical prediction software then defines the transformation function between the actual part profile — the preliminary sample — and its actual cavity profile — the preliminary cavity. Once this transformation function is defined, the desired part profile replaces the first molded sample profile as the variable in the transformation function to calculate the final cavity profile. The transformation function is based on a series of trigonometric and polynomial functions.

The initial cavity profile coordinates are:
M1 = K x D
where D is the desired gear profile coordinates and K is the shrinkage (scale) factor for particular plastic material.

At the same time, the initial cavity profile coordinates can be presented as:
M1 = ƒ(P1) where P1 is the preliminary sample gear profile coordinates and ƒ is the transformation function describing relations between the initial cavity and the preliminary sample gear.

Then the final cavity profile coordinates are:
M2 = ƒ(D)

The comparison is taken and then extrapolated. There are not any fudge factors; the method is based purely on the inspection results and math.

With the genetic molding solution method, controlling every step of the process is important. However, knowing material, mold, or molding process parameters isn't necessary to find the transformation function, because the initial molded part shape already contains this information. If these parameters are changed, the molded part shape is changed and a different transformation function is generated. In short, the genetic molding method completely separates the first molding-process development/optimization stage from the second stage — achievement of the desired part configuration. These two stages are equally critical.

The specifics

To reiterate, stage one encompasses molding process development and optimization. This includes the following four steps:

  1. Target gear data definition

    This is the first data set generated and used later. The computer model of the gear provides the coordinates of the nominal gear profile. The distance between points depends on the size of the gear and the number of teeth; it is typically about 0.002 in.

  2. Preliminary cavity designed profile definition

    The preliminary mold cavity profile is defined as the target gear profile that is simply scaled up by the linear mold shrinkage coefficient provided by the material supplier.

  3. Preliminary cavity actual profile definition

    This is the second data set recorded. Manufacturing and coordinate measurement machine inspection of the preliminary cavity provides the point coordinates for the second data set.

  4. Molding process optimization

    Gears are molded using the preliminary cavity, without concern about the gear shape. The goal here is to achieve a stable and repeatable molding process. Once this goal is reached, the molding process must be locked in and certified; no changes to the process are now allowed. Using the optimized process, several dozen gears are then molded.

    Again, stage two of the genetic molding solution is to fully define the final mold cavity. This includes the following steps:

  5. Preliminary gear sample selection

    All of the molded gears are roll tested and the data analyzed. Then the most representative preliminary gear sample is selected.

  6. Preliminary gear profile definition

    This is the third data set. A coordinate-measurement-machine is used to inspect this representative sample; its geometry is recorded.

  7. Final cavity profile definition

    This is the final output data set. The genetic-molding-solution software uses the preliminary gear sample and preliminary cavity data — the third and second data sets — to generate a transformation function. The target gear data (which was the first data set) is then used as the variable in this transformation function to define the final cavity profile — in other words, the output data set. The final cavity is then manufactured and given a coordinate-measurement-machine check-inspection.

  8. Final gear profile

    At last, gears are molded using the final mold cavity. The coordinate-measurement-machine inspection data of the molded gears should be identical to the specified gear profile, within the molding process accuracy variation.

For more information, call (651)308-8899 or email ak@akgears.com.

One example

The green gear shown is an example of the genetic molding solution put into practice. The gear presented is not particularly molding friendly: It has a metal over-molded shaft, two cams, six spokes, and three injecting gates located in the middle of these spokes. Mold development for this gear using traditional methods requires considerable time and guesswork — not to mention several mold cavities. On the other hand, the genetic molding solution method develops the desired cavity in a very short time by direct calculation, with zero guesswork.

The chart here shows a comparison of roll test graphs on the most representative gear. Plotted in red is the trace of a gear produced by a preliminary mold cavity simply scaled up in size; this resulted in an AGMA class-Q5 gear. Plotted in green is the roll-test plot for a similar gear, but from a mold cavity calculated by our genetic molding method. It is an AGMA class-Q9 gear.