Prototyping: Adding Armor
October 1, 2006
Polymer parts produced by some of the more commonly used rapid prototyping machines are great for evaluating form and fit, but their ability to perform well in functional testing depends upon whether the particular resin can withstand the physical stresses of the specific function. One approach for addressing that concern is to clad the plastic prototype with a thin layer of metal.
Pitney Bowes, Shelton, Conn., the well-known maker of mailing equipment and document handling machines, played an important role in the development of this solution. Like many companies, Pitney Bowes prefers to validate part designs before making a costly investment in tooling. Production versions of critical parts are frequently made of stainless steel or high-performance engineering resins because of the stresses they must endure over repeated use.
The company has long used stereolithography machines to produce prototype parts, but found that the SL prototypes lacked the strength and wear resistance to withstand the rigors of their newer test programs. Moisture absorption was another issue, because that affects dimensional stability of the prototype, and dimensional stability is a key concern for testing moving parts with close tolerances that would be exposed to moisture under testing conditions that simulate normal operation. The company wanted a solution that delivered more durable prototypes, while still preserving the dimensional accuracies achieved by the SL prototyping process.
With that quest in mind, Pitney Bowes turned to RePliForm, Baltimore, which had been using electroforming processes to deposit metal onto SL patterns. Could a thin metal coating adhered to an SL model dramatically improve the model's properties?
According to Sean Wise, president of RePliForm, even a 0.002-in. thick layer can significantly enhance prototype performance because the coating makes the prototype 100 times stiffer and more than 10 times stronger than general-purpose SL resins. The company's tests demonstrated that the cladding of SL models increased impact strength by more than six times and doubled the temperature at which heat deflection occurs.
Pitney Bowes discovered that the copper/nickel-plating technology helped their SL parts meet six important criteria:
- Ability to withstand impact and vibration from product testing.
- Excellent wear resistance.
- Dimensional stability.
- Water resistance without leaking.
- No warping from contact with water.
- Improved aging effects.
The success of the initial project has led Pitney Bowes to frequently employ the cladding technology for functional testing of prototypes.
OriginsRePliForm distinguishes between the terms electroplating and electroforming, the latter which it uses on RP models. Standard plating baths are optimized for finish or hardness, and they require a good bond to the substrate for integrity. By contrast, the acid copper and nickel sulfamate baths used in electroforming are optimized to provide strong, low-stress deposits that will produce freestanding structural elements.
RePliForm had already been using electroforming to make prototype tooling, so there was considerable experience with the process when approached with the Pitney Bowes mission to try it on RP models.
There were several initial concerns. One was keeping the coating thin enough to preserve the functional dimensions of the model. Another was whether such a thin coating could provide the desired stiffness levels. It was also important that the process itself not change the part dimensions by exposure to either high temperature or moisture absorption.
Before electroforming metal onto the parts, the parts had to be made conductive. It was imperative that this step be performed at room temperature without exposing the parts to conditions that would cause the SL models to swell, and it must allow a contiguous metal film to form quickly over the part surface.
Initially, a conductive-silver-filled paint was applied to the SL parts to make them conductive, but the paint layer added to part dimensions and the bond to the coating was weak. Subsequently, it was found that a 1 micron electroless nickel layer could be applied reliably to the surface of the SLA models at room temperature. Even though this process is done in aqueous solutions, the low temperatures did not result in swelling except for models with very thin walls that were made from moisture sensitive resins. After this step, the parts are electroplated with copper, then subsequently plated with nickel.
The initial research focused on SL resins (7100, 9100 and 10100) from DSM Somos, Elgin, Ill., partly because these were already in use at Pitney Bowes and also because these resins exhibit high moisture resistance.
The research began with the preparation of both coated and uncoated test samples made with different resins. The samples were first tested for water absorption and its effect on dimensions. Then they were tested for mechanical properties, including stiffness and creep.
The next series of tests involved metal-coated prototypes of parts used in a Pitney Bowes machine, including a reversing arm. The production version of this part is typically made from a glass-reinforced polycarbonate that has a stiffness of 5,500 MPa (800,000 psi). The reversing arm is a critical, high-precision part because there are a number of moving mechanisms attached to it and because it must endure several hundred thousand cycles over its life.
In the engineering test phase for this part, an uncoated SL prototype typically failed before 10 percent of the required test cycles were completed. A coated version of the prototype part was then prepared. To compensate for any concerns about accuracy, part surfaces were offset inward by 0.05 mm, the intended coating thickness. The part was coated with a base layer of copper followed by a topcoat of nickel.
The metal clad prototype lasted more than 400,000 cycles and was used in the engineering test phase for more than eight weeks while the production tool was fabricated.
Expanding potentialSubsequent research by RePliForm, FineLine Prototyping, a service bureau in Raleigh, N.C., and DSM Somos showed that the electroformed coatings applied over ceramic-filled SL resins can also be used to prototype thin-walled metal parts such as die-castings, and even sheet metal parts, which typically cannot be accomplished with uncoated RP models.
Die-castings with thin walls and fine features must often be prototyped by expensive processes such as CNC machining or the creation of prototype tooling. Tests of sample parts demonstrated that electroforming metal onto a resin RP part be a faster and less expensive means to deliver prototypes for functional testing.
The research into this area characterized the volume of metal deposited necessary to achieve mechanical properties that would simulate typically used die-casting materials. For example, a metal volume fraction of 0.25 delivers a level of stiffness comparable to magnesium in tension, but equivalent bending stiffness could be obtained at a metal coating volume fraction of less than 0.17 to simulate magnesium and less than 0.25 to simulate aluminum due to skin stiffening.
(To determine the coating thickness needed for a specific volume fraction, take the average wall thickness of the parts, multiply that by the volume fraction desired and divide by two because the metal is applied to both sides. So for a volume fraction of 0.2 on a part with average wall thickness of 1 mm, the coating needs to be 0.1 mm thick. To keep dimensions under control, offset the wall thickness on the CAD model inwards by the coating thickness prior to building the model.)
This study also characterized the type of part designs that would be suitable. Experimental parts included a bushing for an armature winding and parts with tapped holes into which threaded fasteners could be applied. FineLine and RePliForm were also able to impress a manufacturer of air compressors by demonstrating the capability of producing metal-clad models durable enough to test the performance of a highly stressed, dynamic assembly. In this case, the assembly consisted of a crankcase, crankshaft, connecting rod, wrist pin, piston, cylinder, and cylinder head, all made with a composite SL resin and then plated to a metal thickness of about 0.18 mm. Using the assembly of coated SL prototypes, the unit was powered up and generated 410 kPa (60 psi) of air.
ConsiderationsWhile much of the development for electroforming RP parts involved stereolithography and DSM Somos resins, the process can also be used on other methods of prototyping and with different resins, though each technology and resin has properties that may make it more or less optimal for a given project.
Part geometry is also a consideration, because the rate of metal deposition is dependent on field strength, which is a function of geometry. Protruding features exhibit high field strength and plate rapidly, while recessed features have low field strength and plate slowly. As a result, coating thickness drops off rapidly in deep, narrow sections. If a recess is large enough, however, a supplementary anode can be placed inside to improve metal distribution, assuming the passage is not oddly shaped. Other techniques can be employed, such as the application of field control elements, that can help to minimize thickness gradients.
Nominal thicknesses for the process range between 0.05 mm to 0.12 mm. The optimal coating thickness must be determined by balancing the need for greater strength against dimensional accuracy requirements.
The process is not suitable for parts that require repeated flexing, such as snap fits, because flexing can cause delamination of the cladding from the substrate. The process is also inappropriate for parts that must act as electrical insulators.
Other applicationsWhile enhancing the mechanical properties of RP models has been the primary focus of the technology, the high electrical conductivity of the metal coatings also permits their use in electromagnetic shielding or reflectivity applications. A thickness of less than 0.01 mm is more than enough to provide good signal attenuation. Applying a coating thickness of 0.05 mm to both sides yields an attenuation performance comparable to a fabricated metal box.
One of RePliForm's customers is a maker of advanced radar systems that previously machined its radar reflectors out of aluminum because it seemed to be a convenient method for producing a precise, low-volume part. But after RF testing a metal-coated SL reflector, the company discovered it could get equivalent performance without the high cost of machining. With volumes as low as five, the company realized savings of 70 percent on the part.
Cost and turnaroundThe cost for metal coating RP models is a function of dimensions, surface area, and desired coating thickness, but as a general rule, the cost of coating a model is approximately 25 percent to 50 percent of the model's commercial price, with small parts pricing at the high end and large parts pricing at the low end of the range.
Turnaround time is typically two to three days for smaller parts requiring less than 0.075 mm of metal applied. RePliForm only coats the parts, so the prototypes must be fabricated and delivered by the customer or its service bureau.
For more information email: sean@RePliFormInc.com
Sidebar: Metallic MakeoverCladding a polymer prototype significantly improves its physical properties. Though the degree of the gains depend upon the base resin, coating thickness, and part geometry, testing by RePliForm has led to some general observations. Samples measuring 127 mm x 12.7 mm x 3.2 mm were used for the tests below.
- Moisture stability:
- Metal coating improves the moisture stability of models by acting as a barrier. Metal-coated samples displayed no dimensional changes after soaking in warm water for three weeks. By comparison, uncoated samples increased in length by nearly 2 percent for DSM Somos 10100 and by 1 percent for DSM Somos 7100.
- Mechanical reinforcement:
- Metal coating significantly improves the stiffness of a model and moderately improves its strength. Bending stiffness of coated samples is up to 10 times higher than uncoated samples when coating thickness exceeds 0.1 mm. Bending strength increases by factors ranging from two to six. (The stiffening and strengthening factors are even higher with thick coatings, but it is more appropriate to think of this in terms of metal volume fraction rather than coating thickness.)
- Creep performance.
- Improvement in creep performance is one of the more attractive benefits of metal cladding. Samples coated with 0.05 mm of metal showed less than 0.1 percent deflection with a constant 10 MPa stress in a three-point bending configuration, and exhibited slightly more than 0.2 percent deflection for 20 MPa stress at one hour. By comparison, uncoated samples showed deflection of 0.45 percent at 10 MPa and 1.45 percent at 20 Mpa. At 500 hours, the coated samples exhibited strains of 0.11 percent at 10 MPa and 0.35 percent at 20 MPa, while the uncoated samples showed strains of 2.4 percent at 10 MPa and 7 percent at 20 MPa. These results demonstrate the greater dimensional stability that can be achieved by cladding.