Compressive stresses in induction treatment?

[Fred]

We manufacture a part approximately 3 mm thick (obtaine by fine blanking) with 3 zones surface hardened (high frequency treatment) for years.

We thought that this kind of treatment lead gave always a compressive state of the heat treated zones until we carried out stress measurements because of micro-crack problems.

We are now very surprise by the results : the presence of a layer of tension stress of about 10 µm depth.

Could anyone bring us his experience.... Thanks

 

[Carburize]

Fred - as we discussed yesterday on carburize case depths - the induction hardening process is designed to cause the surface layer to go through a quench cycle to produce a layer of martensite. The transformation to martensite is accompanied by a volume increase which is resisted by the base material below the transformed layer and this puts the surface layer into compression. The compressive residual stress field at the surface has to be balanced by a tensile residual stress. Typically the stress plots I have seen have a short range surface compressive stress localized close to the hardened surface layer and then a longer range, lower level tensile field balancing the compressive layer.

 

[Fred]

I agree with your point... That's why we are very suprised by the stress field : a tensile field (+200MPa) close to the surface 0-5µm and a low compressive (-50 MPa) fied underneath.

 

[Carburize]

Are the three zones all induction hardened at the same time?
How was the residual stress field measured?

 

[Goahead]

Is it possible that the outermost layer be at lower microhardness than material further inside? Is the treatment based on self quenching or on external quench?
Can a difference in microstructure of the present material be observed vs. material of the previously successful process?

http://www.welding-advisers.com/

 

[TVP]

Fred,

Several things have been mentioned that are pertinent, but it is still necessary to know more about the details of the shape/geometry, and exactly how the part is induction heated and quenched. One thing to keep in mind is that the compressive residual stress pattern is produced by carburizing, but not always by induction hardening. This depends on the geometry and thickness distribution, the nature of the quench flow and impingement on the part, and how the 3 different hardening zones interact. Quite a complex problem. Are you working with your induction vendor and quench supplier?

 

[Fred]

I'm going to try to respond to your questions ....

The 3 zones are heat treated in 2 times but these zones are sufficiently distant to avoid any problem of thermal affected zone.

The material of the part is 65Cr3. And the treatment consists in a high frequency treatment with an inductor in the shape of 2 pins (heating from the faces), quenched with water mixed with a polymer. We use a delayed quench (1.8s)to minimize cracks problems

The part is tempered at 150°C.

The microhardness of the outermost layer is higher than the layer further inside and the microstructure is fine martensite.

 

[Carburize]

Fred - one question you did not answer - how was the stress pattern measured?

 

[Fred]

The stress pattern is measured by diffraction and the matter is removed by electrolytic method from the surface to 50µm in a radius where we have sometimes problems of cracks.

 

[Carburize]

Check the surface for any changes in alloy particularly any evidence of decarburization. The martensite start and finish temperature are directly dependent on carbon content and a 0.1% loss of carbon at the surface can raise the Ms temperature by around 50 degree C depending on the other other alloying elements.
If you have any loss of alloying elements at the surface sufficient to produce a change in Ms and Mf temperatures a thin surface can transform before the main part of the case. When the main part of the case goes through its transformation it will cause a tensile residual stress to be developed in the thin previously transformed surface layer as well as in the substrate material.

 

[Fred]

No decarburation can be seen on the micrography, only a fine analysis ("microsonde") could put a partial decarburation in evidence.

In fact, I beleived that the high frequency treatment was short enough to avoid any decarburation problems in depth.

But to complete the description of our problem, I have to mentionned the "toad skin" of the heat treated zone (without affected zone under skin)

 

[Fred]

I would like to complete this thread with new measurements of the residual stress : the values are more likely up to 700-800 MPa in tension in the direction of the thickness (on an approximately 70 µm depth) all along the martensitic zone whereas we have -500MPa on the faces.

It is interesting to note also a high compresive residual stress before the induction hardening corresponding of the fine-blanking operation (-800 MPa). Could this compressive residual field be responsible for the tension stress by repartition phenomenon?

On other point which could be very important: the part is heated from the faces and cooled perpendiculary by a double-shower ==> the angle between the heated-zone and the spray holes is about 90°. We are principaly working on this parameter and the % of the cooling polymer