Fermilab Helical Spring Failure Analysis

Sloan Digital Sky Survey Telescope Technical Note 19980512 

Packer Engineering, Inc.
1950 N. Washington Street, P.O. Box 353
Naperville, IL 60566-0353
Phone (630) 505-5722 - Fax (630) 505-1986


I. Introduction

Packer Engineering was requested to determine the cause of failure of a small helical stainless steel spring that attaches to a hook on the instrument latch mechanism. Failure occurred after 10,195 warm (18 C) cycles and 523 cold (-18C) cycles. The extension spring is wound from 0.049" wire, and is approximately 2" long and 0.3" in diameter. Figure 1 contains a photograph of the as-received spring.

 

II. Procedure

After visual examination under 7.5 - 64X binocular scope, pertinent areas of the spring surface were visually examined by scanning electron microscope (SEM). This included the two fracture faces, the surfaces adjacent to the fracture faces, surfaces slightly away from the fracture faces and the similar areas at the other end of the spring. Energy dispersive spectroscopy (EDS) was utilized to determine the chemical species present in various locations of interest on the spring surfaces, and the hardness was determined by Knoop microhardness procedures.

 

III. Results

The hardness of the spring was found to be between 47 and 49 HRc. This is considered normal for stainless steel spring. The hardness measurements are contained in Table 1.

Rachet lines and fatigue striations were found on the fracture surface originating at the inside diameter of the spring end loop. The striations are very closely spaced, which indicates that the fatigue crack propagated over a large number of cycles. There were no usage induced defects such as a gouge present at the crack origin. Figures 2 and 3 show the fracture surface and the fatigue striations found.

However, a roughened area was found to be present on the spring wire surface where the fatigue crack initiated. EDS elemental analysis of this area was compared to the surrounding area and the similar area on the other end of the spring. Abundant aluminum, calcium, phosphorous oxide and sulfide contamination's, along with silicon and carbon was present in the roughened area which was not present elsewhere. Figures 4 and 5 show a visual comparison of the roughened area to the adjacent area, and Figures 6 and 7 contain EDS spectra showing the contamination present.

 

IV. Conclusion

The type of contaminants present suggest that their source was during the wire manufacturing process, and not during spring manufacture. The type of compounds found are typical of slag and refractories used in the steel making process. As the steel billet is rolled and drawn, unwanted and normally not present internal contamination (non-metallics) tend to work their way to the surface of the final product form. The subsequent roughness and gouges in that area created stress risers that led to crack initiation under cyclic service conditions.

The scope of work for this project was to perform a failure analysis of the subject spring. Packer Engineering's mechanical and materials engineers have experience in product and process improvement methods. Since the helical spring investigated is of high value in use, we would welcome the opportunity to specify a proper solution to the spring failure.

Clarifications and questions regarding this report can be handled by either writer by calling (800) 323-0114

 

 

PACKER ENGINEERING, INC.

 

   

Steve Strack

Frederick E. Schmidt, Jr., Ph.D., P.E.

Manager

Director

Materials Engineering

Materials Engineering

 

 

 

 

 

 

 

Table 1

 

MICRO HARDNESS TEST

   

EQUIPMENT

Leitz Wetzlar Microhardness Tester

TEST BLOCK I.D. #

MPA 3704872

INDENTOR WEIGHT

300 g.

HARDNESS

HK0.1-624Kp/mm2

BEFORE

625Kp

AFTER

624Kp

PART / MICRO I.D.

Micro # 4830

 

 

           

Load

Location

Measured

Corrected

KHN

Hardness

 

Length

Length

 

HRC

1

edge

88.8

91.2

513.22

48.6

2

edge

88.2

90.6

520.04

49.0

3

edge

90.0

92.4

499.97

47.7

4

middle

88.8

91.2

513.22

48.6

5

middle

88.7

91.1

514.34

48.6