PWB Fabrication Cost and Energy/Mass Balance Modeling

[Revo & Evo Programs] | [Material-Centric Model] | [Illustrative Example] | [References]

The "Revo" & "Evo" Programs

A consortium of industry fabricators, equipment manufacturers, and suppliers is working together to develop new photopolymer products and processes to reduce the cost of building printing wiring boards and make the fabrication process more friendly toward the environment. For a process to be attractive and thus acceptable to industrial fabricators the new processes must have a lower overall fabrication cost, and if possible, a performance enhancement. The fact that the materials and processes reduce the waste load and toxicity of materials that must be managed or sent to the environment provides an additional incentive. Two projects, funded in part by the Defense Advanced Research Projects Agency (DARPA) under contract to USAF Wright Laboratory through the Environmentally Conscious Electronics Manufacturing Initiative, take both an evolutionary and a revolutionary approach to changing the manufacturing process.

Material-Centric PWB Fabrication Model

In traditional activity-based manufacturing cost models, activities are based on equipment and facilities ("equipment-centric"). In this methodology the process steps are based on material processing activities ("material-centric"). Equipment-centric models are appropriate for IC manufacture where the processing cost is driven by facilities and equipment, however, in printed wiring board (PWB) manufacturing, where a significant portion of the cost is materials, it is more appropriate to focus the process modeling around material activities.

In a material-centric PWB fabrication model, each activity or process step is defined in terms of what it does to the materials associated with the substrate being fabricated. Fundamentally, five material activities are used in the model:

In addition, activities 1-4 may have associated consumable materials. Consumables are materials that are attached to the process (as opposed to the product). Consumables are used (and wasted) by the activities, but, at no point in the process do consumables reside in the product. Examples of consumables are: water, artwork, and drill bits.

While process steps that model the above material activities contain information about the equipment and facilities required, the process step is not defined by the equipment and facilities. By defining process steps in terms of their material treatment, modeling of material and waste costs becomes straightforward.

The material-centric model was implemented in an existing activity-based cost analysis implemented in the SavanSys tradeoff analysis tool, enhanced with material and energy inventorying. The cost model resides in an existing multidisciplinary tradeoff analysis tool for multichip packaging tradeoff analysis. The basic process flow modeling available within this tool is summarized in the Figure.

Figure 1 - Process step model used in the SavanSys tradeoff analysis tool for material-centric PWB cost modeling.

During the execution of a process flow, inventories of material in the product, material in the waste stream, and energy consumed are created and manipulated. Each material inventory catalogs material volume at Standard Temperature and Pressure (STP) and the material's name. As each process step is executed its material and energy requirements are computed and added to or subtracted from the respective inventories. Some activities transfer materials between inventories, e.g., if the step produces waste materials by removing material from the product, the quantity of waste is subtracted from the material used inventory and added to the material waste inventory. All the inventories are normalized to one instance of the part being processed, i.e., the inventories keep the used and wasted materials that correspond to a single panel or board - if the total waste is desired, the contents of the waste inventory must be multiplied by the number of panels or boards processed.

Illustrative Example

This section presents a simple illustrative example analysis performed using the material-centric cost model.  The example presented here is the fabrication of a double-sided, un-drilled, Copper clad layer-pair for use in conventional PWB fabrication. The costs as a function of process step are shown in Figure 2. The plot shows cost broken down by labor, material (computed using the methodology discussed in this paper), tooling, capital, and lumped. The "lumped" cost represents the effective cost of yield loss. The total cost of an 18x24 inch layer-pair fabrication in this case is $19.38 with a yield of 94% per layer-pair (yielded cost = $20.62/layer-pair). Approximately half the cost of the layer-pair is the cost of the laminate inserted in the first step. Other significant contributors include the resist (steps 6 and 7), artwork (steps 9 and 10), and the AOI inspection step at the end of the process flow. The AOI inspection step includes the cost of performing the inspection (labor and capital) plus the reallocation of money spent on layer-pairs that are scrapped by the AOI into the layer-pairs that are passed by the inspection.

Figure 2  - Cost as a function of process step for the fabrication of a double-sided, undrilled, copper clad layer-pair. The embedded pie chart shows the fractional distribution of labor, material, tooling, and capital costs. The portion of the material cost that ends up being wasted is shown as a raised pie slice.

The pie chart included in Figure 2 shows the relative contributions of labor, material, capital, tooling, and yield loss to the cost of an layer-pair. The pie chart also shows the fraction of the cost that has been invested in material that is wasted prior to completing the processing of the layer-pair (waste disposition associated with the waste material from layer-pair fabrication is not considered in this example). The AOI step includes material in addition to labor and capital contributions, while the AOI activity has no direct material cost associated with it, it is scrapping layer-pairs that contain material investments and the portion of those material investments that are allocated back into the cost of the layer-pairs that are passed by the AOI is shown as a material cost associated with the AOI step.

The materials used corresponding to the layer-pair fabrication are shown in Figure 3. Nearly all the material that is present in the final layer-pair is added by the Copper clad laminate (first three steps). Addition of the resist is shown in steps 6 and 7. The develop process removes all of the resist except that which covers the metal features desired on the layer. Etching removes all the Copper that is not protected by the resist, and strip removes the remainder of the resist. There are other second order variations in the material usage that are too small to be seen in Figure 3 such as a slight reduction in the volume of material in the layer-pair in the "Mylar Removal" steps in the middle of the flow. Mylar removal is the step where Mylar that is protecting the resist layers is removed and discarded.

Figure 3 - Material used (material in the product) as a function of the process steps associated with fabricating a double-sided, un-drilled, copper clad layer-pair.

Figure 4 shows the material wasted as a function of process step in the layer-pair fabrication. Most of the waste generated is water from the rinse activities, however, nearly 1000 cc of non-water waste is also produced. The non-water waste is primarily composed of developer and stripper but also includes cleaners, resist, Mylar, and artwork. Note that the first three steps (the initial laminate insertion) produces no waste for the layer-pair fabrication. Obviously once the layer-pairs are used in the fabrication of an actual PWB, there will be considerable waste laminate generated when the individual boards are singulated from the panel. That waste laminate will not be inserted into the waste inventory until the singulation activity occurs, i.e., all the laminate used to make the layer-pair is still part of the layer-pair at the end of this example. The AOI inspection step at the end of the flow contributes significantly to the waste inventory because it scraps some of the layer-pairs that have been produced. All the materials in the scrapped layer-pairs and all the waste allocated to the scrapped layer-pairs must be reallocated into the waste inventory for the non-scrapped layer-pairs. Included in the waste that must be reallocated is the waste water used to produce the scrapped layer-pairs as well as the original laminate and resist used to produce the scrapped layer-pairs.

Figure 4 - Material wasted as a function of the process steps associated with fabricating a double-sided, un-drilled, copper clad layer-pair.

References

P. A. Sandborn and C. F. Murphy, "Material-Centric Modeling of PWB Fabrication: An Economic and Environmental Comparison of Conventional and Photovia Board Fabrication Processes," IEEE Trans. on Components, Packaging, and Manufacturing Technology – Part C, vol. 21, April 1998, pp. 97-110.

P. Sandborn, "Integrating Environmental Inventory Analysis with Detailed Printed Circuit Board Fabrication Cost Modeling," in Proc. 19th AESP/EPA Pollution Prevention & Control Conference, Jan. 1998, pp. 177-183.

P. Sandborn, "Performing Design Tradeoffs for Advanced PWBs," IPC National Conference on Organic High Density Interconnect Structures, Santa Clara, CA, pp. 131-149, Nov. 1997.

P. A. Sandborn and C. F. Murphy, "Evaluating the Cost Impact of Design-for-Environment Decisions Early in the Product Design Cycle," Proc. of the IPC Works, Washington DC, pp. S03-9-1 to S03-9-7, Oct. 1997.

C. Murphy, P. Sandborn, and J. Lott, "An Environmental and Economic Comparison of Additive and Subtractive Approaches for Printed Wiring Board Fabrication," Proc. IEEE International Symposium on Electronics and the Environment, May 1997.

P. A. Sandborn, J. W. Lott, and C. F. Murphy, "Material-Centric Process Flow Modeling of PWB Fabrication and Waste Disposal," Proc. IPC Printed Circuits Expo., pp. S10-4-1 - S10-4-12, March 1997.

C. F. Murphy and P. A. Sandborn, "Cost and Environmental Metrics for a High Density Laminate Technology," Proceedings of the Sixteenth International Electronics Manufacturing Technology Symposium (IEMT), Austin, TX, Oct. 1996.

P. A. Sandborn and G. McFall, "Performing Design for Environment Concurrent with Interdisciplinary Tradeoff Analysis of Electronic Systems," Proc. IEEE International Symposium on Electronics and the Environment, pp. 167-172, May 1996.