Guidelines for Rapid Prototyping Success

Terry T. Wohlers

A variation of this was published in the November 1991 issue of Rapid Prototyping Report by CAD/CAM Publishing
Copyright 1991 by Wohlers Associates

Rapid prototyping (PR) doesn't just happen. It takes a tremendous amount of research and planning. If you make good decisions when evaluating, purchasing, and installing an RP system, you are likely to experience success -- success you will want to share with the world. But if you ignore issues such as training, personnel, and the role of the CAD system, your experiences may be terribly disappointing and maybe even disastrous.

RP pays back when applied in several areas. It can be difficult to justify and pay back an RP system if it is used for only one purpose. For example, building concept models are unquestionably beneficial early in the design process, but their savings in time and dollars are impossible to quantify. Consider using the system for other purposes too, such as soft tooling, investment casting and fit and function testing.

Consider RP for soft tooling. In most cases, a pattern for a mold can be produced faster with an RP system than using conventional fabrication techniques. Quick turn around of patterns means faster turn around of molded prototype parts. This shortens the design cycle and reduces cost.

Use RP to communicate with others. Concurrent engineering means getting feedback from others early in the design process when changes can be made inexpensively. Manufacturing, marketing, purchasing, customers, and other groups contribute early in the design. RP can play a strategic role in facilitating concurrent engineering because parts can be produced quickly and made available for review. When implemented properly, it can improve the design of a part, save manufacturing costs, and reduce the time it takes to bring a product to market.

Don't underestimate the role of the CAD system. Drawings that come out of a printer or plotter are only as good as the CAD drawing file. The same is true with rapid prototyping. Poor CAD models produce poor RP parts. Unfortunately, most CAD work being done today is 2D drafting and wireframe modeling, and neither will work as input to an RP system. A fully closed surface or solid model is required. Companies new to surface or solid modeling may spend as much or more time developing their CAD modeling skills as they will with the RP system.

It may require two or more tries to make good parts. Placement of support structures (for those RP systems that require them) and part orientation is not always obvious at first. Also, many factors must be considered such as shrinkage, cure depth, and draw speed. Understanding how the system reacts when changing these parameters requires lots of practice and experimentation with the system.

Part accuracy depends largely on user experience. Accuracy of 3D Systems' SLA-250, for example, currently stands in the +/- 0.002" to +/- 0.005" per inch range, but accuracy can vary widely depending on the part geometry and the skill of the user. New users without a good grasp of shrinkage factors and the many operating parameters of an RP system typically do not make parts as accurate as experienced users. Also note that while certain dimensions of a part may fall well within a +/- 0.005" or less per inch, other dimensions on the same part may be much worse. Experienced users can better predict the outcome before the part is built.

RP is not best for all parts. RP permits you to produce incredibly complex parts and this is where the technology really shines. But it may not be practical to use the resources of an RP system to produce basic parts when conventional tools and machines are less costly to use, better understood, often more accurate, and offer a much wider selection of materials.

Budget for the CAD installation. Set aside $15,000 - $45,000 for the CAD system. Mature surface or solid modeling software will range from $5000 to more than $20,000, depending on the specific options you select. The graphics workstation with adequate MIPS, MFLOPS, memory, and disk storage will cost $10,000 - $25,000. Allow for training and weeks of practice before expecting results.

Budget plenty for the RP system. Commercially available RP systems from 3D Systems, Cubital, DTM Corp., Helisys, Light Sculpting, Quadrax Laser Technologies, and Stratasys cost between $85,000 and $490,000. Helisys represents the lowest cost option while Cubital offers the highest. Prices have not dropped much since stereolithography was introduced more than three years ago, but the price/performance ratio, fortunately, has improved. For instance, systems available today are much faster and produce parts more accurately with better materials and surface qualities. Prices are expected to drop in the future as developers improve existing systems and introduce new RP techniques such as ballistic particle manufacturing.

Consider RP system maintenance. Normally, RP system maintenance agreements cost $16,000 - $36,000, but the actual maintenance costs can vary widely from system to system. RP systems without a Helium Cadmium or Argon laser will generally cost less to operate and maintain. The cost to replace an SLA-250 laser is $9,200. The SLA-500 laser is priced at $25,000. Cubital's Solider system does not involve a laser but the complexity of the system may require special attention and replacement parts.

Consider the initial cost of material. A mile-long spool of wax or nylon filament for Stratasys' 3D-Modeler system is priced at $350. A gallon of photopolymer is priced at $300 - $550. The SLA-250's 7.8 gallon vat costs $2,730 to fill at $350 per gallon. Budget $23,450 to fill the SLA-500's whopping 67-gallon vat.

Your people are key. Individuals with a can do attitude, coupled with enthusiasm, will carry the project forward. People that constantly find reasons why something cannot be done, will delay and maybe even bring the project to a dead stop. Be careful when choosing the people that investigate, implement, and manage the technology. They can make the difference between success and failure.

Find a champion. Identify one individual to champion RP technology. This person must be progressive thinker and willing to take a risk. The individual must view RP as an opportunity for the organization, as well as for him/herself, but must be cautious. The individual must understand that if good decisions are made, the outcome will be positive, but that good decisions won't come without lots of hard work.

Keep an eye on the future. CAD/CAM system developers Auto-Trol, Computervision, GE Calma, Control Data Corp., Gerber, Sperry Univac, and others pioneered the CAD/CAM industry. Some of these companies continue to do well, but many no longer exist or are struggling. Some of their products were the best at the time, but more practical options have become available.

The same may hold true in the RP industry. Many researchers are developing new RP technologies at universities, government agencies, and in the private sector. In fact, 27 different organizations (including those named above) are researching and developing technologies that are unique in some way. Certain developers believe that large build chambers coupled with better materials and improved accuracy is the answer. Others believe they can cash in on the estimated 1.5 million desktop CAD system users by offering them a low-cost desktop unit.

Copyright 1991 by Wohlers Associates