Printer Friendly

Evaluating boss designs for thermoset glass-filled polyesters.

Among other factors, the study examined the effects of fiber orientation and distribution and their relationship to the failure of the boss.

Thermoset glass-filled polyesters are widely used as engineering compounds in applications such as structural components for computers and other electrical devices that require high dielectric strength. In the design of such parts, proper assembly of components is a major concern, and the increased use of thermosets is creating a need for a better and cheaper means of fastening. The common approach to joining two plastic pieces is the use of a screw-boss mechanism, whereby a boss is molded into one piece and joined to another piece by a screw.

Proper design of the boss is essential. If the boss is undersized, stresses resulting from insertion of the screw and the load placed on it could cause the boss area to crack and fail. If the boss is oversized, problems in manufacturing will follow. Oversized screw bosses lead to an unnecessary increase in cycle time and to cosmetic problems, such as sinks and voids resulting from thick walls.

Properly choosing a screw is also important. How often the screw will be taken out and reinserted is among the factors that must be considered. Also, the design must take into account the different screw configurations, which create varying amounts of stress on the boss. Many different designs for screw bosses can be found in reference materials.(1-3)

This article evaluates some of these designs. The goal is to show how well the boss performs under certain tests, and to determine if the boss design is over-engineered or lacking in strength. The difference in screw designs as they affect performance is also evaluated.

Theory/Background

It is a general rule in the design of screw bosses that the boss be removed from the outer wall of the part. The boss can be connected to a wall by a thin rib of approximately 40% to 80% of the nominal wall thickness. However, use of the thin rib is dependent on material and cosmetic requirements, because when the boss adjoins a wall, it can create a thick section that leads to sinks and voids. Additionally, the hole in the boss can result in an undesirable weld on the outer cosmetic walls.

A general guideline for boss design is that the inside diameter - commonly the pitch diameter of the screw - usually depends upon recommendations provided by the screw manufacturer. If a thread-cutting screw is used, a space underneath the screw will be needed for chip relief. If this area is not provided, chips from the formation of threads will build up and lead to more stress in the boss area.

The outside diameter of the boss is the most widely discussed in boss design. Outside diameter guidelines range from 1.5 to 3.5 times the inside diameter. Thicker bosses better distribute the stresses, thereby reducing the tendency to crack. However, increased wall thickness increases cycle time and cosmetic problems, such as sinks and voids. It is therefore very important to choose the proper outside diameter. The boss usually has some radius on both the inside and outside edges of its base. A commonly accepted value for the outside radius is 0.25 times the nominal wall.

Two main categories of self-tapping screws - thread-cutting and thread-forming - are used with plastics. Thread-forming screws deform the plastic into which they are inserted. Thread-cutting screws physically remove the material in the same way that a machine tap removes material to form a thread path.

Thread-forming screws are usually less expensive than thread-cutting screws and produce the highest resistance to backout. However, they cause a high hoop stress in the part. Thread-forming screws are generally used with lower-modulus plastic material because ductility or cold flow is a prerequisite for their use. The screws are generally used with plastics that have a flexural modulus of less than 2.76 x [10.sup.9] N/[m.sup.2]. They are used for more ductile semicrystalline polymers or toughened thermoplastics, where residual stresses can relax to an acceptable level. Thread-forming screws are also suitable for applications where limited re-insertion is anticipated.

Thread-cutting screws are most commonly used with higher-modulus thermosetting or thermoplastic polymers that do not have the ductility required for thread-forming screws. Hoop stresses brought on by insertion are relatively low. Thread-cutting screws can be used in situations where disassembly or assembly is anticipated. However, the strength of the boss will deteriorate rapidly if additional threads are cut during reassembly. The primary goal in design of screw bosses is to meet the stress requirement of the screw and use the least amount of material. Any decrease in material used will reduce production costs. Besides lowering cost by requiring less material, thinner walls decrease cure times and increase production. For these reasons, proper design of the screw boss in an assembly must be carefully calculated.

Failure in a screw/boss assembly can occur in several different ways. The boss may split during insertion because of high insertion stress, the screw may pull out of the boss under tensile load, or the boss may separate from the nominal wall under tensile load. Two specially designed jigs were built for this study to evaluate the various modes of failure.

Procedure

A screw boss was designed according to guidelines set by a company that manufactures engineered self-tapping screws. The inside diameter measured 3.4 mm and the outside diameter 11.6 mm. The boss was 12.825 mm high, with a 0.254-mm radius at the point where the boss meets the nominal wall. The boss was attached to a plate measuring 50.8 x 101.6 x 3.1 mm. To allow for proper chip relief for a thread-cutting screw, the screw hole was 28.7 mm deep. The ratio for inside to outside diameter was roughly 3 to 1. A transfer mold was built to produce this part, and the parts were produced using a glass-filled polyester thermoset resin.

Two different screw designs were used in the study. One is an "engineered" screw - a thread-cutting screw specifically designed for thermoset and thermoplastic resins. The screw has a 30 [degrees] thread angle, narrower thread profile, and recessed thread root, which lowers drive torque by providing additional space between threads for material displaced during cutting. The other screw is a generic, self-tapping screw with a thread angle of 60 [degrees]. This type of screw is used for many other applications, including assemblies with metal.

Both screws were inserted into two different boss designs with the same 12 mm of thread contact. Two different test jigs were designed and used to fixture the boss to the lower jaw. The first fixture contained an oversized hole through which the boss could freely pass, and restrained the nominal wall to which the boss was attached. Unlike most screw test jigs, the fixture provided for evaluation of potential failure of the boss at its junction to the nominal wall. The second test jig constrained the boss at its top surface, limiting failure to screw pullout. The top jaw held the head of the screw. A tensile load was applied at the crosshead speed of 50.8 mm/min, and the pullout strength of each screw was evaluated. The test was repeated ten times for each screw.

The failure of the boss was observed, and the outside diameter was changed to produce a 2-to-1 ratio between inside and outside diameter. The mold was changed and the redesigned bosses were produced under the same conditions as before.

The screws were inserted into the new design and again tested on the Instron machine.

Results

The parts were tested in the jig holding the plaque at the nominal wall. In all the tests, the bosses failed in the same manner, cracking about 2 mm up from the plaque on the gate side of the boss. The crack then proceeded around the boss, heading downward and eventually pulling out some of the nominal wall. Neither the engineered nor the generic screws pulled out of the boss. Specimens were then tested in the second test jig for pullout strength. The engineered screw had a pullout strength of 195.220 kg, and the generic screw a pullout strength of 172.520 kg.

Following completion of this test, the boss design that had an outside diameter twice that of its inside diameter was evaluated. In both screws, the bosses cracked lengthwise upon insertion of the screw. The generic screws cracked the boss quickly upon insertion; the engineered screw took a longer time to crack the boss and required less torque to insert. Because of the failure during screw insertion, no further tests could be performed for this design.

After the specimens were tested, they were examined to determine how the fibers were oriented upon molding, and their relation to the mode of failure. As Fig. 1 shows, the fibers were aligned perpendicular to flow between the gate and boss. As the melt hit the boss, the fibers bunched up at the base of the boss and did not fill out the top of the boss with the same concentration [ILLUSTRATION FOR FIGURE 2 OMITTED]. In addition, the transverse to flow orientation was maintained, resulting in orientation of fibers perpendicular to the length of the boss. As the melt moved around the boss, the fibers began to orient themselves parallel to the gate; this orientation continued into the nominal wall. In regions of the part where flow was not disrupted by the boss, the fibers continued to align themselves perpendicular to the gate as the part filled out [ILLUSTRATION FOR FIGURE 1 OMITTED].

Discussion

All the bosses failed in the same manner. They cracked on the gate side, radially around the boss and then in a downward direction, eventually pulling away some of the nominal wall. Inspection of the failure revealed that the fibers were bunched up at the base of the boss near the gate. A lower number, it seemed, made their way to the top of the boss. The failure originated near the base at the region of high fiber concentration. The crack then proceeded to move parallel with the orientation of the fibers around the boss and headed into the nominal wall.

The boss was designed relatively well. The small radius at the base seemed to be a possible point of crack initiation, but was found to be unimportant because the failure occurred above it. An outside diameter that measures three times the inside diameter is a useful guideline. With a decrease in outside diameter, the bosses were failing during screw insertion because of excessive hoop stress placed on them by the screw. The radius that was placed on the base of the inside hole seemed adequate because the crack did not separate the radius.

Orientation of the fibers had the biggest effect on how the boss failed. The crack originated at a weak point where fibers were bunched up perpendicular to the direction of tensile force. Because of the high concentration of fiber, this region would also be resin-poor. Because the fibers in this fiber-rich region were oriented perpendicular to flow, they would provide no reinforcement. This, combined with the low concentration of resin, resulted in a particularly weak region. As the crack developed in this region, it followed the direction parallel to the fibers around the boss and down toward the nominal wall.

The engineered screws did not outperform the generic screw in the pullout strength test. However, the engineered screw required less torque and created less hoop stress during insertion.

Conclusion

The boss design provided by the screw manufacturer performed reasonably well. However, the manufacturer's recommendations did not address the factors that would limit boss failure. Strength of the boss was limited most by how the fibers were oriented. The crack originated at a point where fibers were oriented perpendicular to the force acting on the piece. The boss created a problem with how the fibers were oriented throughout the part. Fibers were oriented perpendicular to flow; then, when they entered the boss region, they changed orientation to become aligned with the direction of flow. This leads to areas on the boss and nominal wall where drastic changes in orientation of fibers occur, reducing strength.

The engineered screws did not outperform the generic screws in the pullout test. It was observed, however, that the engineered screws required much less torque during insertion in the boss. When inserted into the smaller boss, they had much less hoop stress.

Although the engineered screw provided more pullout strength than the generic screw, failure of the boss was not limited by this factor. Failure resulted from boss separation from the nominal wall, not stripout force. Therefore, the lower insertion torque and hoop stress provide a positive feature, while promoting a significant pullout strength averaging 195.220 kg.

Future work in this area will include work on boss design to enhance proper orientation of fibers so that they align parallel to the pullout force placed on them. Such an orientation will improve the overall strength of the boss.

Acknowledgments

The author wishes to thank Ed Fish, Ed Fish Machine Co.; Chuck Lang, Penn State Erie, The Behrend College, The Plastics Technical Center; and John P. Beaumont, Penn State Erie, faculty advisor.

References

1. J. Burton, "Advances in Self-Tapping Screws," Plastics Design Forum, Nov./Dec. 1990, pp. 54-58.

2. K.J. Gomes and J.F. Braden, Mechanical Fasteners Used in Plastics, Marcel Dekker, New York (1984).

3. C.T. Keller, "Self-Tapping Screws: How to Choose and Use the Right One," Plastics Design Forum, Nov./Dec. 1983, pp. 56-58.
COPYRIGHT 1996 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hill, Wiliam
Publication:Plastics Engineering
Date:Jan 1, 1996
Words:2271
Previous Article:Overcoming process control challenges of gas-assist injection molding.
Next Article:Designing interphases for "mega-coupled" polypropylene composites.
Topics:

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters