Rapid Prototyping Cost | Rapid Prototyping Industry

Justifying rapid prototyping

Sorovetz, Thomas

Rapid prototyping allows both designers and manufacturing engineers to see and hold physical models of part designs on demand--often in under a day after the prototyping process begins. Models can be used to visualize designs; verify engineering changes; check for form, fit, and function; and verify part producibility. Sometimes, they can be used in test applications. Resulting cost savings over conventional methods can be significant.

Justifying the necessary equipment can be difficult, though. One reason is that key benefits of rapid prototyping aren't always easy to quantify. For instance, how do you quantify the benefits of increased communication between design and manufacturing engineers] Another is that the financial staff may not understand the need for a new technology that can radically change the way a company produces prototypes. Here's how I prepared a successful acquisition proposal.

Demand Rise

Several rapid prototyping technologies are available today. The method chosen depends on the application. For example, the stereolithography apparatus (SLA) from 3-D Systems (Valencia, CA) excels at building plastic models with thin walls and intricate internal structures. Another method may be more efficient at large, thick-walled structures.

SLA is a layer manufacturing process. Users must develop a CAD solid model or fully closed 3-D surface model of both the part geometry and support structures required to produce the prototype and convert these to an STL file format. A "slice" computer automatically divides the STL file into cross sections. These slice programs drive a laser across a bath of photopolymer (liquid plastic), successively solidifying each layer to produce the model. After postcuring the plastic model can be finished by sanding, sandblasting, painting, or dyeing and used for various applications, including conceptual design, prototype parts, and patterns for metal castings.

Chrysler Corp. began using rapid prototyping (RP) in 1990 when Jeep/Truck Engineering acquired two new SLA machines and support equipment. Our goal at the Rapid Prototyping Lab was to provide RP services at no cost to all engineering groups within Chrysler to help reduce tooling costs and rime to market.

By 1992, the SLA equipment was in almost continuous use seven days a week, supporting T300 and PL vehicle programs. Maximum capacity was about 550 geometries (models) and 1300 parts/yr. Fed by model data designed with Chrysler's Catia CAD/CAM systems, it produced a typical prototype part in about 13 hr.

Results after two years of rapid prototyping were promising. The average plastic model saved about 320 man-hours and $10,500 in out-of-pocket costs compared to traditional methods. Although difficult to extrapolate, in-house SLA savings were probably in the range of $5-$10 million. The RP Lab also was operating at maximum capacity.

It was time to ask management for new equipment. Justification would be based on the need to support new vehicle programs and the need to benefit from software advances that would increase the number and types of parts that could be prototyped on an SLA. We expected the capital acquisition process to flow a little smoother than the first time, since management was now familiar with rapid prototyping and Chrysler had documented several success stories. Still, the proposal had to avoid supplying reams of data without sufficient explanation. It also had to be in a clear, concise format that both financial staff and engineering management would understand.

Basics of the Capital Proposal

A well-written capital equipment acquisition proposal has several key ingredients, including the following-:

*The Executive Summary. This should be no more than 1/2 to one page in length and describe an operation's current status and why the new equipment is needed. It also can be beneficial to include a productivity improvement factor for the proposed acquisition. This is equal to the total savings from the acquisition over a five-year period divided by the cost of the acquisition and support expenses throughout five years of use. We've found that a successful capital project usually has a productivity improvement factor no less than four and no greater than 10. With this RP equipment acquisition, for instance, we anticipated a fourfold increase in savings by year end 1993.

*The wish list. Ours included two SLA-250 machines, two Silicon Graphics Iris "slice" computers, postcuring apparatus, and two laser units. This equipment, along with upgrades to existing equipment requested in a separate proposal, would double our RP capability. We evaluated both new and used equipment and provided a total cost for each option, including costs for equipment transportation and installation, installation support, and systems integration support.

During our initial foray into rapid prototyping, we bought new equipment from the vendor for several reasons. The first was knowing that the equipment would be supported by that vendor. In addition, training would be included in the acquisition package, and maintenance provided for the first year. With used equipment, you have no idea where that machine's been or how it's been maintained. Moreover, you may need to replace elements, such as the laser and power supply. You also have to pay a fee to have the technology certified for maintenance by the original vendor. Nevertheless, once you have some experience with the technology and know what to look for in used equipment, there are some good deals out there. This is why this proposal was essentially a request to buy used equipment.

*Alternatives. A well-written capital project also details the alternatives to the acquisition and their possible downsides. For example, one alternative in this RP project was simply to upgrade existing equipment by buying a new computer workstation (already part of another proposal). The problem was that this alternative would increase productivity 25-30, but would be insufficient to address the fourfold increase expected in demand for services.

Another option was to use an outside service bureau. For small companies, this may be the best route when exploring use of rapid prototyping. One year, Chrysler averaged more than $300,000 in service bureau business. Cost of this proposed RP equipment acquisition was a little over $200,000, so it made more sense to have the equipment in-house. Moreover, with service bureaus, RP equipment use was subject to limited availability, and the bureaus we examined couldn't always meet required turnaround times.

*Case histories. In any capital acquisition request, you should do your homework. Find out how your competitors are using the technology, and document their published results. Examine the technology's impact on production flexibility, responsiveness to market changes, quality and reliability improvements, human resources, inventory levels, even customer satisfaction.

For our first RP proposal, we faced the problem that our people didn't understand the technology. We solved that by including a videotape of rapid prototyping along with documented case histories from competitors. For our second proposal, we had several proven Chrysler applications of rapid prototyping, including the following:

1. Experimentation on an intake manifold SLA model allowed us to determine that the flow test of the SLA model was equal to that with the conventional prototype. We saved three iterations of prototypes by testing and revising the stereolithography models of the intake manifold.

2. An "A" pillar blocker couldn't be tooled in time to support the F1 prototype build. Using stereolithography, we created a model from surface data and made a tool off of that to create a mold. This saved some $30,000 in design and CAD time, iteration of the A" pillar blocker design, and production of a hand mold.

3. A model for a cylinder head flow box typically takes 320 hr (2 months) to fabricate at a cost of $10,000. Rapid prototyping produced the model in two weeks (80 hr).

Early in Chrysler's rapid prototyping history, we faced another problem that many new users will experience, that of convincing the design people that they now had to start designing in solid models or "volumes" so we could create the STL file to drive the system. This brings home a key issue in any acquisition project for rapid prototyping. You may start out focusing on the technology's immediate impact in your area, but don't be surprised if there is a ripple effect--the technology has potential to improve efficiency of both upstream and downstream applications.

Copyright Society of Manufacturing Engineers Dec 1995
Provided by ProQuest Information and Learning Company. All rights Reserved