Truly Functional Testing: Part 1
How to select RP materials so that valid conclusions about the production part can be drawn from testing RP parts.
There has been remarkable progress made in
the development of materials for RP processes over the last few years.
Early RP materials were so brittle and fragile that they were useful
for little more than concept models. New materials, however, have
increased the functionality of RP parts significantly, so much in some
cases that RP parts are being used for limited production.
Material Introductions
Today’s RP materials more closely
approximate injection-molded plastics. In some cases, thermoplastic
materials are actually used in the RP process. The laser sintering
processes, both DTM and EOS versions, have long used nylon materials as
their workhorse materials. Both offer unfilled and glass-filled
versions of the material. The fused deposition process from Stratasys
also uses thermoplastic materials including ABS, polycarbonate, and
most recently, polyphenylsulfone.
The nature of stereolithography (SL)-type processes precludes the use
of thermoplastic materials directly. However, in the last few years, a
wide range of new resins have been introduced, some of which are
claimed to have similar properties to common injection-molded plastics,
including ABS, polypropylene, and polyethylene. DSM Somos has even
introduced filled SL resins with stiffness in the range of glass-filled
plastics.
There is no question that the introduction of these new materials has
significantly increased the value and utility of RP systems to
developers of injection-molded plastic components. RP systems are
capable of creating more than concept models; it is possible to create
RP parts that can be used in functional tests. That means that
designers can evaluate several design iterations in less time and for
less money than they used to spend to evaluate just a few. The result
is better products getting to market faster, at less cost.Limits
As valuable as the introduction of more functional RP materials has
been, there are limits to how useful they are. Despite the improved
functionality of the materials, designers still cannot answer these
questions:
- If the RP prototype passes testing, does that mean the production part will pass the same test?
- If the RP prototype fails testing, does that mean the production part will fail the same test?
Those questions cannot be answered, because the RP material is not the same as the production plastic. As good as RP materials have become, they do not behave exactly as the injection-molded plastic they claim to approximate. Even thermoplastic RP materials such as the FDM and SLS differ significantly in material properties from the molded plastic parts, due to variations in material specifications and differences in processing. They simply will not duplicate the properties of a molded material.
Consequently, designers have not been able to draw conclusions about the adequacy of a design based on testing RP parts. In order to ensure that the design is adequate, it has typically been necessary to build a tool and test parts that were molded in the production material.
If it were possible to build RP parts in a material that behaved exactly like the production plastic, it would also be possible to verify design adequacy by testing RP prototypes, eliminating the need for molded test parts. Designers could go right into production tooling with the confidence that the design was adequate.
None of the currently available RP systems, however, will allow parts to be built in a material that behaves exactly like a production plastic, nor will any of the systems in development.
Testing
It is usually not necessary, however, to match all the material properties of the production material. In nearly all testing situations, the performance of the material is determined by relatively few material properties. For example, consider a common plastic coat hanger. The ability of the hanger to do its job—support the weight of a coat—is dependent on two material properties; stiffness, which determines how much the hanger will deflect under load, and strength, which determines how much load it can withstand before failing. Other properties like heat deflection temperature, density, dielectric constant, etc., are largely irrelevant. Consequently, if we could build a prototype hanger using a material that matches the stiffness and strength of the plastic material proposed for production, we could test the prototype and draw conclusions about design adequacy without having to mold a part in the production material.Note that this would only verify adequacy of the design for the static load situation. If the hanger also needed to withstand an impact load, different material properties would be involved and testing the prototype developed for the static load situation may not provide useful information.
Consequently, if we can select a material which matches the key material properties, the RP part will predict the performance of the production component. The RP part can be used for design verification testing, avoiding the cost of molding test parts.
Selecting an Appropriate RP Material
How can we select an RP material that will adequately approximate the production plastic? The process is straightforward and consists of four steps:Step 1: Identify the prototyping situation. Clearly define the requirements for one aspect of design adequacy. For example, in our hanger example, it might be to support the static load of a winter coat. This may not be the only performance requirement for the design.
Step 2: Determine the key material properties for the prototyping situation. In each situation, relatively few material properties will control the performance of the material in that situation. By examining the prototyping situation closely, it can be determined what material properties control the performance of the material in that situation.
Step 3: Determine how closely those material properties must be approximated. It will be impossible to match material properties exactly in every case. However, if the properties of the RP material are close enough to those of the proposed plastic, it is still possible to draw valid conclusions. The challenge is to determine how close it is necessary to be for a valid conclusions to be drawn. Once determined, a range of values will be specified for each key material property. Any RP material that falls within that range can then be used to build a prototype valid for adequacy tests.
Step 4: Select an RP material and process. Now that the bounds of the range of material property values are known, it is relatively simple to search the available RP materials, determine which are within that range and select one for the prototype build.
To illustrate this process, four common prototyping situations are examined.
Prototyping Situation 1:
Rated Load Mechanical OperationStep 1: Identify the prototyping situation. This prototyping situation is probably the most common design adequacy testing done on mechanical parts. The goal is to verify that the design will perform adequately in normal operation, within rated load, and not be subject to unusual conditions. It is not intended to verify maximum load or life, only that it will perform as intended within the rated load of the device.
Step 2: Determine the key material properties. For this situation, two material properties will be key. The first is a measure of stiffness. Stiffness determines how much the component will deform under load and how rigid a component feels. There are several measures of stiffness to choose from, but flexural modulus is most appropriate for plastic parts, becauseplastic parts are most often stressed in flexure rather than tension or compression. A wide range of stiffness is now available in RP materials. In the last few years that range has been extended significantly with the introduction of filled SL resins.
The second property is some measure of the strength of the material. The most appropriate is flexural strength, again because plastic components are most often stressed in flexure.
Step 3: Determine how closely to approximate the properties of the intended production material. For flex modulus, the prototyping material should be within 20 percent of the flex modulus of the production material. Since variations in the stiffness will not affect the stress, only the amount of deflection in the component, the performance of the material is not critically dependent on the exact value of stiffness.
Flexural strength of the prototyping material does not actually have to match the production material, only be high enough that the part will not fail at rated load. Nearly all parts are designed with a factor of safety so that temporary overloads will not fail the part. Since we defined the prototyping situation as being within rated load, we do not need to consider overloads in the selection of the material. The factor of safety is typically two or higher. Consequently, if the flexural strength of the prototyping material is greater than half the flexural strength of the production plastic, the prototype will not fail during testing.
The approximation requirements have now been defined. The prototyping material must have a flexural modulus within plus or minus 20 percent of that of the proposed production plastic and a flexural strength greater than half that of the proposed production plastic. Any RP material that falls within those bounds will adequately simulate the production material. If a part made from such a material survives testing, there is a high level of confidence that the production part will also pass the same test, even before parts are molded in that material. If the RP part fails during testing, there needs to be concern about the production part.
Step 4: Select a rapid prototyping material. The final step is to determine if any RP materials fall within the bounds defined by the approximation requirements.
Assume that the material of choice for production is an ABS material, specifically Cycolac MG47 from GE Plastics (a common ABS). MG47 is plotted as a black point. Surrounding that point is a rectangle defined by the approximation requirements discussed above; plus or minus 20 percent of the flexural stiffness and greater than one-half the flexural strength. Any RP material that falls within this rectangle will adequately simulate MG47 for this prototyping situation.
Also plotted are many of the currently available RP materials. Material properties for RP materials were obtained from data published by the manufacturers (with the exception of the polycarbonate and ABS materials from Stratasys, where samples were purchased from a service bureau and sent to a materials lab for testing). If the manufacturer’s published data listed a range of values, the midpoint of the range is used in this analysis. A number of manufacturers do not list all material properties for their materials. In those cases where data is not provided, the materials are not included in the chart. For example, flexural strength information is not published for selective laser sintering (SLS) materials, preventing their inclusion on this chart.
Several materials fall outside that rectangle and therefore would not be suitable. It is interesting to note that the ABS material from Stratasys does not adequately simulate the MG47 material, illustrating that differences in processing can result in significant differences in material properties.
There are a number of materials, however, that fall within the rectangle. They include the 11120 and 14120 stereolithography resins from DSM, the polycarbonate and polyphenylsulfone materials from Stratasys, the SI40 and AccuGen resins from 3D Systems, and the 7560,7565,7580 and 7510 stereolithography resins from Huntsman. Consequently, a prototype constructed from any of these materials will adequately simulate MG47 in this prototyping situation and allow valid conclusions to be drawn about that design.
We can make a similar chart for any injection-molded plastic and quickly determine which RP materials, if any, can adequately simulate that plastic for functional prototyping so that valid conclusions can be drawn about the performance of that material in that application.
Because most service providers stock a limited number of RP materials, and their customers collectively use dozens, if not hundreds, of plastic materials in production, it can be more useful to present this information in a slightly different format. Instead of a rectangle around a plastic material defining the range of RP materials that can adequately represent it, we can build a rectangle around an RP material that defines the range of plastic properties it can simulate. The five DSM Somos resins used at Express Pattern are plotted in black. Around each is a rectangle that defines the range of plastic properties it can adequately simulate, based on the same approximation requirements discussed above. Several common plastics for injection molding are plotted on the chart. The plastic materials are color coded by type with the key at the right side of the chart. This is now a quick reference chart to show which resin should be used to simulate a plastic for this prototyping situation.
With this straightforward process, one can quickly determine, for any prototyping situation, which, if any, RP materials have properties close enough to those of the target plastic that an RP component can be used in verification testing for any prototyping situation. The situation discussed, rated load mechanical operation, is just one of many potential prototyping situations that might be encountered in product development. In the second article in this seroes, three other prototyping situations will be examined, and guidelines for using this process will be presented.
For more information, contact Tom Mueller of Express Pattern (Buffalo Grove, IL) visit the Web site at www.expresspattern.com
