Using PC-based seal analysis reduces prototypes.
Geauga does rubber extrusions and rubber molding as well as plastic extrusions, plastic injection molding and plastic blow molding. The Middlefield, OH facility limits its work to rubber, producing dual durometer products that are made by combining two different kinds of rubber to form one cross-section. Bulb seals, a type of seal used extensively on automobiles, are a typical dual durometer product produced by the company. These seals can be found inside the doors, around the trunk, under the hood and around windows. Bulb seals can also be found in appliances such as dishwashers and washing machines as well as overhead doors.
A dual-durometer bulb seal consists of a bulb made of a softer material and a harder rubber U channel that grips to the sheet metal of the car. The purpose of having the soft bulb is to permit a closure without a great deal of effort but one that seals against water, outside materials or wind noise. Typically, the customer requests a certain closure force and contact pressure for a seal that Geauga must meet when it manufactures the seal.
In the past, the design process for a bulb seal began with hand drawings showing the product in place, on a car door for example. The hand drawings were re-created as CAD drawings depicting the sheet metal body, the, sheet metal door and the seal itself After that, it was necessary to find the right combination of materials so the seal would meet customer's specification for closure force and contact pressure. This was done through a lengthy build and test cycle and there was no way to know how a seal would perform until an actual seal was made.
This required the production of an extrusion die that closely matched the CAD drawing. This step may take many weeks and cost several thousands of dollars depending on its complexity. These figures include the cost of the material and the processing. They also include time spent by the engineers and operators and the time equipment was taken away from production.
Once the die was ready, about 45 to 60 minutes were needed to set up extrusion lines and run the product. If the die needed modification, that took from 15 minutes to several hours. For a new seal, it usually took several iterations to get the die right. Once the seal was produced, it was subjected to tests that simulated the car door closing and calculated how much effort it took. The seal was set up on the test fixture, the door was closed and the forces needed for closing were recorded. This could take many hours, and anywhere from one to 10 tests were needed to obtain a seal with the right degree of closure force.
The time required to create a die, extrude rubber through it and perform physical testing added significantly to the time needed to produce a new product. This led the company to look into simulating seal performance on a computer. Most commercially available packages that are capable of handling the non-linearity of the seal materials ran on either high-end workstations or mainframes, bringing the cost of a typical analysis to several thousand dollars. Geauga chose the personal computer-based package EASi-SEAL from EASi Engineering, Bingham Farms, MI, because it made seal analysis affordable and was the only software designed specifically to address the critical issues of seal design. The company purchased version 1.2 of EASi-SEAL, which supports the design of multi-durometer seals.
The design process for the new car door bulb seal began as it did in the past, by modeling the geometry of the seal and the surrounding sheet metal in the CAD system, either EDS Unigraphics or Autodesk's Autocad. That data was then imported into EASi-SEAL as an IGES file. The engineer supplied the program with preliminary information, identifying which component was the seal, the body and the door, then specifying parameters such as how far the door should close into the seal and the speed of the closure. It was also necessary to specify material data in the form of a stress-strain curve. The software stores material information in a database, so once the properties of this material were entered, they can be recalled for future analyses. EASi-SEAL also allows the user to combine up to five different materials in the same seal, making it possible for Geauga to simulate this dual durometer rubber seal accurately.
Error checks on both material and geometry, such as whether all the required material data have been given or the closure impinged on the seal, were performed by the software automatically. Next, EASi-SEAL automatically meshed the seal, extruded the geometry into solids, applied boundary conditions and material properties and performed a finite element analysis that took both material and geometric non-linearities of the problem into account. This analysis, which ran without user intervention, took between one and two hours.
For the engineers, the most important output from the analysis was the closure force for the seal. They used that information to modify the original design, then repeated the analysis and modification process until they got the force they needed. Because computer simulation was so much faster than physical testing, it was possible to investigate how the new bulb seal's closure force was affected by different combinations of variables. For example, engineers simulated the effect on the seal of two or three mm variations in build, combined with 10 point variations in rubber hardness and variations in the wall thickness of extrusion. They ran analyses at nominal values and at the worst-case values in the high and low directions. This gave the engineers and the customer insight into how the seal would perform in worst-case production scenarios, information that was not available in the past because samples were run at nominal size and shape. This information ultimately helped the customer improve quality by alerting them to possible problems.
Other useful output from the software was a frame-by-frame animation showing the seal cross-section as it deformed. This was particularly helpful to the seal designers because it gave them a theoretical understanding of seal performance that they had never been able to get prior to physical testing. The program provided other information as well, including CLD (compression load deflection) curves, stress/strain contours and pressure distributions.
The primary advantage of this approach was the time and money it saved by reducing the number of build-and-test iterations needed to design the new seal. The design they ended up with as a result of the computer simulation was close to the final design. Some modification was needed when the actual seal was produced, primarily for appearance reasons, but only two prototypes were needed.
The company has so far used EASi-SEAL in the design of five new products. For the first few products, engineers verified the EASi-SEAL results against physical testing. They found an average correlation of 91%. This high level of correlation was maintained over many iterations as shown by comparing relative relationships in the different dimensions. This has given the company confidence in this design approach and what it can offer to customers. The Geauga Co. now creates better initial designs and needs significantly less lead time to get to the final product.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||personal computer|
|Date:||Mar 1, 1996|
|Previous Article:||Record consumption for rubber in 1995.|
|Next Article:||Advances in silicon engine gasket sealing.|