jchumley@columbus.rr.com

ABSTRACT: A method is presented which models the failure mechanism of a fully evacuated pre-stressed television picture tube protected by conventional shrink banding when impacted by either a 51 millimeter diameter hardened steel ball or a 2.3 kg missile projected at the face of the tube. Underwriter’s Laboratory, Inc. standard for safety UL1418 cathode-ray tubes describes this test. The finite element model uses a combination of explicit and implicit codes to account for the contact forces and preexisting stress in a brittle viscoelastic material (glass). A combination of high speed photographs, strain gauge measurements and fractography is used to verify the model. The method has obvious commercial significance because previously there has not been any reliable method to predict impact behavior of a CRT tube until the actual product was available for testing.

                                      Figure 1 test setup
 

INTRODUCTION: In December of 1992 Underwriters Laboratories, Inc. (UL), in conjunction with the Canadian Standards Association issued the fifth edition of the standard for cathode-ray tube (CRT) safety UL1418. The intent of the standard was to inform the public of the potential for injury in the event that a tube fails to meet the standard. However, Underwriters Laboratories, Inc. does not assume any liability to anyone for use of or reliance on the standard. To describe the test briefly, a CRT is mounted in a test cabinet as shown in Figure 1.  And struck with either a 51 millimeter diameter, hardened (60 Rockwell C), steel ball or a 2.3 kg missile. The impact areas on the tubes and the initial kinetic energy of the projectiles are described in the standard. The standard specifically requires that in the case of ball test: a) no glass particle having a mass of 0.025 grams or greater be shall be thrown between the 150 millimeter high barriers located 0.9 and 1.5 meters from the test cabinet. b) The total mass of all pieces of glass between the 0.90 and 1.50 meter barriers shall not exceed 0.10 grams. c) There shall be no glass expelled beyond the 1.50 meter barrier. For the missile impact test: a) There shall be no single particle having a mass greater than 14.2 grams between the 0.90 and 1.50 meter barriers. b) The total mass of all particles between the 0.90 and 1.50 meter barriers shall not exceed 42.5 grams. c) There shall be no single piece having a mass greater than 1.4 grams beyond the 1.50 meter barrier. Over the years CRT manufactures have developed a variety of strategies to protect the tube, but the current method most widely used is to pre-stress the tube with a shrink-fit steel band around the perimeter of the tube see Figure 2.

                            Figure 2 Shrink-Banded tube with sag-back recovery Gauge.

While it may be difficult to predict the final disposition of all the fracture fragments the present work concentrates on modeling the static pre-loading of the tube (using the ANSYS^TM finite element code) and the dynamic impact event (using LSDyna^TM code).

DESCRIPTION OF THE MODEL: The entire tube is modeled using hexagonal elements see Figure 3. A pressure load of 1 atmosphere 101,353 Pascals is applied to the exterior surface of the tube. The shrink band is modeled 0.05% larger than the circumference of the tube and contact calculated using a thermal contraction of 0.1%.

This methodology insures that the band is stretched beyond its yield point. Since the ANSYS ^TM model is used as the starting point for the LSDyna ^TM analysis of the impact event it was not possible to take advantage of the quarter symmetry in the ANSYS ^TM model. Also, since the LSDyna ^TM code requires an element with only 8 nodes it was necessary to further refine the mesh size to assure the accuracy of the results. In the case of the ball impact model it was only necessary to replace the band elements with the equivalent contact pressures and rerun the static case then run the explicit solution.

                     Figure 4 Ball impact model elements

For the missile test two additional refinements were required. The UL standard requires the face of the CRT, at the top and the bottom shall be scratched 3 mm from the screen or phosphor edge into the viewing area. The scratches shall be horizontal lines, 100 mm long. Because it would not be practical to model the scratches in detail the elastic modulus of the effected elements was modified to simulate a stress concentration factor of 3.0.

The second requirement is that the impact object travel shall be restricted so that the rounded end of the missile penetrates the CRT face, equal to or less than 25 milliliters. The missile is constrained by attaching a cable element to the end of the missile.

Solving the equations of motion of the missile for the above given boundary conditions

yields an effective cross-sectional area of the cable.

RESULTS: The results of the ball impact model were not very impressive since, as often happens in this test, the ball did not break the glass (see Figure 6 and Figure 6a).

The standard prohibits increasing the impact energy therefore, it would be necessary to redesign either the CRT or the shrink-band to obtain a different result. That was not done. The missile model provides a richer source of information. With the addition of the stress concentration lines the tube fails dramatically see Figure 7. The failure sequence starts with a horizontal crack across the bottom scribe line followed by a vertical crack origination on the bottom scribe line traveling toward the missile impact location. Next cracks begin to radiate from the impact site starting with a horizontal crack. The fractures starting at the bottom scribe line form from the outside of the tube, while the ones at the impact source form from the inside first.

                                                            Figure 7 Missile Impact Test

 

                                                                    Figure 7a Band Shrinking onto tube Figure 7b Band contact status

VARIFICATION: Several methods were used to verify the modeling techniques beginning with a simple measurement of the so-called spring-back recovery. In this test the change in height of the center of the tube relative to the edges before and after the tube is evacuated is compared to the change after the band is shrink-fitted to the tube. The percent change in relative height recovered with the shrink band is the spring-back recovery (See Figure 2). The model predicted a recovery of 75%. Test data is 66 –87%.

The second method of validation used was high-speed photography. The 20 J missile test was photographed with a photec ^TM rotating prism camera at shutter speeds of up to 10,000 frames per second. Then transferred them to NTS video for analysis. The intent of this test is to capture the progression of crack propagation. While quite impressive, especially in the case of the full implosion, (See Figure 9a) the photographs were not quite fast enough. The speed of sound in the glass is approximately 3,000 millimeters per millisecond. If a crack propagates at 60% of that velocity then a crack could travel 180 millimeters between frames. Nonetheless the general failure sequence described in the results section above was observed.

Figure 9 Missile Impact

Figure 9a imploding CRT just prior to implosion 10,000 Frames/sec

Another test was performed with strain gauges placed strategically on the tube and connected to a high speed recording device also capable of recording 10,000 readings per second on each of five channels. Strain gauge information and fracture analysis of the fragments allows for fine tuning. (See Figure 10).

Figures 11 and 12 show a typical white-light micrograph of the origin of the crack and its direction of travel. From the Walner lines it is apparent that the crack is propagating from right to left and from the mirror radius the stress at the time of failure can be estimated to be between 34 and 41 Mega-Pascals.

FUTURE WORK: As noted in the description section above, the breaking of the tube in the ball test was not predicted. (The hole and cracks shown in Figures 8 and 8a occurred after repeated hits and, as such, are not covered by either the model or the UL standard.) Further study should include a case in which the tube breaks in the ball test, as the failure mechanism is quite different than for the missile test. Also a parametric study of design factors such as glass thickness and shape, shrink-band design is needed to meet the challenges of the new larger and flatter CRT’s being introduced. Also a post-processing technique need to be developed to calculate particle throw.

CONCLUSION: A method has been presented which allows for modeling of an Implosion Protected Cathode Ray tube. With further refinements it should be possible to predict, in advance, those designs which are likely to fail the UL 1418 standard and recommend the appropriate countermeasures.

ANSYS users Guide 5.6 (online-Help System)

Choi, J. H. "Stress Analysis of Safety Band in Glass Bulb Design " SID 1992 Samsung Corning

Frechette, V.D. "Failure Analysis of Brittle Materials" Americn Ceramic Society Westerville, Ohio 1990

Halliday and Resnick "Physics for Students of Science and Engineering" Second Edition John Wiely and Sons, Inc. 1965

Rork and Young "Formulations for Stress and Strain" fifth edition. McGraw-Hill Book Company New-York 1966

Standard for Safety Cathode-Ray Tubes UL 1418 Underwriters Laboratories, Inc. Northbrook, Illinois 1992

Varshneya, Arun K. " Fundamentals of Inorganic Glasses" Academic Press, Inc Boston MA 1994

I would like to thank Rufus Blackley of Ohio Computer Aided Engineering , and Deno Ciabattoni of ANSYS, Inc for invaluable help with the modeling process; Dewey Martin and his staff at Video & Communications Design for arranging the high speed photography; Greg DeuBoise of Columbus Testing Laboratory for doing the stress-strain testing of the shrink band material; Charlie Hartung and Bryan Raynor of the Techneglas Columbus Laboratory for performing the UL-1418 tests and recording the strain gauge measurements; M. Kyono of NEG for general consultation; and Louis Spanoudius for preparation of the micrographs of the fracture fragments.