There are a few international and company specifications for spring materials that have clauses specifying cleanness, or freedom from non-metallic inclusions (sulphides, silicates, aluminas, oxides and titanium carbo-nitrides, mostly). When I speak to steelmaking companies in some parts of the world, they ask me questions about how clean their steel needs to be, especially for suspension springs. My answer surprises most inquirers.
Based upon the premise that I hardly ever diagnose non-metallic particles as having any influence upon spring failures, I assert that cleanness does not matter.
 |
|
Figure 1: Superclean SiCr. |
 |
|
Figure 2: 17/7 PH.
|
There are two notable exceptions to this premise: One is engine valve springs or similar high-performance dynamic springs, and the other is constant force springs. Engine valve springs do fail by inclusion-initiated fatigue. In such instances, the inclusion is always 15 microns or greater, and is usually between 150-450 microns below the shot-peened wire surface.
Engine valve spring specifications might be expected to have clauses that limit the maximum size of inclusions near the surface of rod or wire. Commercial specifications do this, but no international specifications as yet; I am sure they will soon.
Having written this, it was pointed out to me that the American specification for CrSi valve spring wire, ASTM A877 does have a cleanness clause that limits the size of surface and core inclusions. This specification has limits for sulphide, silicate and oxide inclusions, which are most unlikely to affect spring fatigue performance. It also limits alumina inclusions.
It is important to note that heavy alumina inclusions (even occasional ones) are the very inclusions that initiate fatigue failures in engine valve springs.
Spring steel with very few inclusions near the wire surface, and almost none greater than 15 microns, is often called “superclean.” (Absolutely no inclusions is as difficult as zero defects to achieve; keep trying but you won’t ever succeed.) Contrary to what its name implies, superclean wire does have inclusions. However, they are silicate inclusions (see Figure 1), which elongate and are, therefore, below the critical size at which they might harm spring fatigue life.
The other type of springs that fails from inclusion initiated fatigue is constant force springs that are made from cold-rolled pearlitic strip. For reasons that I don’t understand, the raw material for this important class of springs is not covered at all in any national or international specifications. If this material were to be described in a specification, then a clause limiting the size and frequency of non-metallic inclusions would be appropriate.
Apart from engine valve and constant force spring specifications, it is my opinion that there is no need for a cleanness clause until such time as someone can demonstrate conclusively that inclusions have significantly influenced the risk of spring failure due to fatigue, corrosion, stress corrosion cracking or embrittlement. To illustrate this opinion, IST recently tested a batch of 17/7PH springs. Now, 17/7PH has more inclusions than any other grade of spring steel, as shown in Figure 2, page 53. The fatigue, relaxation and corrosion resistance of these springs was very good indeed.
Both wires shown in Figures 1 and 2 meet the criteria of having no inclusions greater than 15 microns, and so the springs made from them gave excellent performance.
High-strength steels frequently have cleanness specifications in order to enhance fracture toughness and impact strength. In these circumstances, cleanness should be specified; but for springs the impact strength does not matter. A correctly designed spring operates only in the elastic range. If it goes plastic, it has failed. Therefore, the impact strength of springs steels, which is often extremely low, need not be specified either.
The moral of this tale is that cleanness clauses are only needed for springs made from valve-quality wire and constant force springs, and then they should be of the type that limit the maximum size of inclusions at positions where the inclusions could do harm – at or just below the strip surface, or within 0.5 mm of the wire surface. Furthermore, impact strength typically doesn’t need to be specified for spring materials.
One final note. “Cleanness” is the term I have used, but sometimes it appears as “cleanliness,” as in “AMS 2301, Aircraft Quality Steel Cleanliness.”
