Most spring steel gets its strength from cold working, heat treating or a combination of both. Typically, small-diameter wire can easily be formed into springs without any problems. However, with wire over 5/8", even the toughest coilers encounter problems when winding spring wire, especially with high-tensile steels.
 |
| Figure 2: Even small alloy additions to steel can retard the transformation of austenite during cooling [2]. |
There are two common options available when coiling large-diameter wire or bar stock: winding in the fully annealed condition or hot winding. In both cases, the springs are typically heat treated after they’re coiled.
Heat treating large-diameter springs, or any piece of steel for that matter, runs the risk of creating distortion and quench cracks. Distortion can occur when non-uniform stresses are present during the quench. Oil quenching produces a slower quench rate than water quenching and, therefore, less distortion. There are many causes for quench cracking, beginning with the surface condition of the steel before heating and ending with the metallurgical conditions during the draw or temper.
 |
| Figure 1: The knee (red line) in a typical transformation curve for steel. Note the inherently slower cooling rate of the bar center vs. the surface [1]. |
For most applications, the main purpose of heat treating steel is to produce a uniform martensitic microstructure. During heat treat, steel is heated up to a point where the entire structure transforms into austenite. When spring steel is heated above ~1500°F, it transforms to austenite. During cooling, austenite is an unstable solid solution of steel. When the cooling rate is fast, austenite transforms into martensite. When the cooling rate is slow, the transformation favors softer microstructures, like bainite, ferrite and pearlite. Martensite is more desirable in the final product because it is significantly harder than pearlite and tougher than bainite.
Avoid the Knee
Any heat treater is familiar with the “knee” in the austenite transformation curve. It can be seen highlighted in red in Figure 1.
 |
| Figure 3: A typical quench crack originates from the surface and progresses toward the core of the bar. |
Why is the knee so important? To assure uniform transformation from austenite to martensite, the steel must cool fast enough so that the cooling profile does not intersect with the knee of the curve.
Whenever the cooling rate of the steel intersects the knee of the curve, the amount of martensite in the final product is sacrificed. It depends on where it intersects the curve and for how long, but in general, slow cooling usually manifests itself as the appearance of acicular bainite in the microstructure, which is softer than the desired martensite. The cooling rate in oil-quenched steel should never be slow enough to form pearlite.
It is important to note in Figure 1 that the center of a bar will never cool as fast as the surface and is always more likely to intersect the knee of the curve. Alloys are added to hot-wound steels to move the knee to the right and enable the entire cross-section to convert to martensite without any deleterious transformations.
 |
| Figure 4: Intergranular fracture surface of a quench crack that was found on a hot wound SAE 5160 spring. Note the characteristic “rock candy” appearance of the surface. The fracture surface was covered with a layer of temper scale. |
For this reason, thick cross-sections typically require greater alloy content than thin ones. Note the effect on the position of the knee after a 0.25% molybdenum addition, as shown in Figure 2, above. To maximize the amount of transformed martensite in large-diameter springs, the type of alloy and rate of cooling should be engineered to avoid any potential intersection with the knee.
As a steel spring transforms from austenite to martensite, the part undergoes a metallurgical expansion, as martensite is less dense than austenite. The surface transforms to martensite before the center. Initially, during the quench, the surface is in compression. Then, as the center transforms to martensite (expands), the surface is in tension. At this point in the process, the bar is very vulnerable to quench cracking. The tensile surface stresses and the brittle martensite have very low toughness, which makes the steel susceptible to fracture until the steel is tempered. At this time, any surface irregularities, inclusions or defects act as stress risers and can initiate quench cracking.
The goal is to quench the steel rapidly yet uniformly. There are several variables that can be controlled in order to produce uniform cooling. A list of the most critical variables is presented below:
- Surface condition before quenching should be free of excessive scale, as this can locally retard the cooling rate. Surface seams, laps or inclusions can facilitate the formation of quench cracks.
- Reheating rates, either during austenitizing or tempering, that are too rapid can cause excessive internal stresses and lead to cracking.
- Quenching from too high a temperature can cause several problems:
- Surface decarburization and grain growth – both of these issues will compromise the surface integrity.
- The formation of retained austenite, which can cause degradation in mechanical properties and dimensional stability.
- Most alloy steel is quenched in oil. Some of the more important variables related to maintaining oil properties are:
- Oil-quench temperature consistency; the oil should not be too hot or too cold.
- Oil cleanliness; as oil is used, the formation of sludge and varnish are promoted, which can cause erratic surface cooling. Excessive sludge can plug filters and contaminate heat-exchanger surfaces.
- Quench oil needs to be properly agitated to foster uniform heat removal.
- The physical properties of the oil should be monitored over time to ensure no degradation in performance. Ideally, control limits should be in place for the following variables:
Viscosity, as it will increase in time.
- Moisture content, which can cause uneven cooling and lead to the formation of soft spots.
- Contaminants, which can be detected by testing the oil’s flash point, acid number and cooling curve analysis.
- Cracking can occur during quenching or anytime before tempering. A good practice is to temper the steel after quenching while it is still warm, or about 150-200°F. Untempered martensite is brittle and contains high residual stresses; it will crack if not tempered soon after quenching. [3]
Features of a Quench Crack
The fracture typically originates from the surface toward the core of the bar in a relatively straight line, as seen in Figure 3.
Quench cracks are typically free of decarburization and contain light temper scale, indicating that they were formed sometime after the quench and before or during the temper. The crack may initiate at a surface irregularity, like a longitudinal seam or lap. When surface indications are present prior to the reheat, they are typically decarburized to some degree. If the quench crack did not originate at a prior existing mark, then it is common to see 45° shear lip at the surface.
The fracture surface of the quench crack is predominantly intergranular in nature, as seen in Figure 4.
After reading this article, some of you are probably glad that you don’t heat treat your springs at all. However, if you do heat treat your springs and have had some issues with quench cracking, maybe this article will help you apply a meaningful solution to the problem.
References:
1. Van Vlack, Lawrence, Elements of Materials Engineering, Editor: Cohen, Morris, Addison-Wesley Publishing Company, 1980, P. 431.
2. Ibid p. 427.
3. Powell, Gordon, ASM International Handbook Committee Chairman, Volume 11, Failure Analysis and Prevention, Editor: Mills, Kathleen, ASM International, 1986. pp. 94-95. v
