Wanted pristine but broken Dilvar head stud for analysis

[reclino]

I want to take a look at and have a materials engineer evaluate one or more failed Dilvar head studs. It might help to have one that has not yet failed. The perfect example would have no other damage or corrosion, but appeared to fail for no good reason.
I would assume many engine builders have a bucket of the things lurking in a corner.
I will post all details of the informal report I get. I hope to be able to have a chemical analysis run at the least so we all know what the heck this stuff is made of. I assume it is a variation on Kovar a Nickel-Iron-cobalt alloy.
I can pay shipping to North Carolina.
I am cross posting this from the 930 forum
David

[Flieger]

I find it interesting that while kovar, invar, etc. are made to have extremely low coefficients of thermal expansion, dilavar is just the opposite, intended to bridge the gap between steel and aluminum. I have no idea what the alloy is.

[reclino]

I had found reference to a low expansion alloy Dilvar P1 before but with more searching I came up with the following.
Carpenter High Expansion "22-3" alloy (ASTM B-753) - with 22 percent nickel, 3 percent chromium and balance iron - has thermal expansion properties higher - equal to or greater than 19.8 x 10-6 per °C (11 x 10-6 per °F) - than any of the alloys in the Type 300 stainless steel series.
Carpenter High Expansion "19-2" alloy (ASTM B-753) - containing 19 percent nickel, 2 percent chromium and balance iron -has a thermal coefficient of expansion similar to that for Carpenter High Expansion "22-3" alloy. The 19-2 alloy, however, can be manufactured to higher tensile strengths for bimetal thermostats that must be stronger and/or springier.

[chris_seven]

Can't really help with a stud but the basic composition of Dilavar is relatively well known.

%Carbon: 0.63 to 0.73
%Silicon: 0.50 Max
%Manganese: 4.50 to 6.00
%Phosphorous: 0.30Max
% Sulphur: 0.30Max
%Chromium: 2.50 to 4.00
%Molybdenum: 0.20Max
%Nickel: 11.00-13.00
%Vanadium: 0.75 to 1.20
%Niobium: 0.10 to 0.40
%Nitrogen: 0.01 to 0.1

Its Werkstoff Number is 1.3937 and it is produced in Germany by Deutsche Edelstahlwerke amongst others.

The combination of Nickel and Manganese at the above levels will stabilise the Austenite Phase at Room Temperature which is why this alloy is non-magnetic.

It is generally used in a Solution Treated and Age Hardened condition with a UTS of 1200MPa and a Coefficient of Expansion (CoE) of approximately 19.4 x 10^-6 m/m/degK.

Elongation is generally good at around 10% and Young's Modulus (E) is typically 198 GPa.

It is safe to describe this material as a Precipitation Hardening Austenitic Stainless Steel.

Its Nickel Content may seem relatively low as most Stainless Steels with 11-13% Nickel tend to be either Ferritic or Martensitic (which also makes them magnetic) but the relatively high content of Manganese has a significant influence in terms of Austenite formation and retention. Unfortunately it has been shown that using high levels of Manganese in relatively low Nickel content steels has an adverse effect on resistance to pitting and crevice formation in corrosive environments.

Austenite, of course, has a Face Centered Cubic Atomic Structure (FCC) and this is the reason for the higher CoE than the more common Ferritic/Martensitic Steels which are Body Centred Cubic (BCC).

It is interesting that Invar is also an Austenitic FCC structure but exhibits a CoE if around 1.3 x 10^-6 m.m/degK which is quite contradictory. Its discovery in the 1920's was of great significance and initially the theory behind its low expansion was based on its confusing magnetic behaviour. Invar is also magnetic in the temperature range where its CoE is low and this is clearly an anomaly. There were attempts made to explain this behaviour in terms of magnetic moments but these proved incorrect and thus far no real explanation has been found but there will be an underlying magnetic reason.

The typical microstructure of a well produced Dilavar is very straightforward and comprises of fine grained Austenite (Niobium is added as a grain refiner) with Chromium Carbide precipitates which strengthen the alloy.

These carbides which should be coherent with the Austenite matrix for maximum strengthening form within the grains of the material but are also found to be present at grain boundaries and this is what gives rise to potential problems.

The basic mechanical properties of Dilavar are encouraging and the alloy should be not only strong but with a 10% elongation should be quite tough and used correctly I am sure that this is the case.

Typical failures of Dilavar studs show a brittle fracture in the direction of the maximum shear stress in the stud (approx 55 degrees to the tensile axis).

One of the issues with Austenitic precipitation hardening stainless steels is that they do have a strong tendency to suffer from Stress Corrosion Cracking (SCC) particularly in the presence of Chlorides. The relatively high level of carbon found in Dilavar does it no favours in this respect.

Stress Corrosion Cracking is a mechanism which generally occurs due to defect initiation at the grain boundaries of a material. The presence of chrome carbides in the microstructure will almost certainly help to create this type of defect. Add to this an increased tendency for the formation of crevices due to manganese then problems can occur.

As the length of the intergranular defect increases it effectively creates a short but growing crack. Once this crack reaches a critical length the preload stress present causes a brittle fracture.

There have been many, many significant catastrophes caused by this mechanism particularly in the oil/gas industries which operate in Sweet/Sour Gas environments.

The resin coating applied to the latest generation of Dilavar Studs appears to have all but eliminated the failures experienced due to this mechanism but the cost of these parts is still relatively high.

There is also apocryphal evidence that Dilavar studs suffer from Hydrogen Embrittlement but I am not sure that this is service related.

It is certainly possible that poor 'pickling' practices used during manufacture could result in failures of this type but there is nothing that would produce this type of failure in the service environment.

Dilavar alloys are now also commonly manufactured in China and some care is needed when sourcing this type of stud.

It would be interesting to test a modern Dilavar Stud for susceptibility to SCC but as a metallurgist with a background in Fracture Mechanics and Fractography I am reasonably confident that the early generation Dilavar studs fail for this reason.

It is fair to say that much of the knowledge we now have with regard to low Nickel Austenitic steels wasn't available in the late Seventies when Dilavar was first introduced.

You also have to question the use of Dilavar on the exhaust side of SC engines and conventional steel on the inlet side. Not really sensible.

[perelet]

I have set of 964 coated studs, plus one broken - i also have high res scan of broken stud, described here:

broken dilavar stud study - Rennlist Discussion Forums



my lo tech understanding - they should be considered wear items and replaced every 10 years. Metal fatigue and cast defects will make them break.

[chris_seven]

Internal rust? I hope you are not being serious?

[perelet]

Had to drop my engine out and take it all apart, because of bloody stud - no joking there.

About rust - I'll let ya'll metallurgy people to tell me why, previous guy said there's iron in the mix, iron will do rust.

[chris_seven]

I agree that broken studs are a PITA and we see them quite commonly but usually on the bottom row of SC engines and many of my colleagues tell me that the latest 993 studs are much better than the early studs.

The fracture surface of your broken stud is unpleasant and does indicate that the material is not in a great condition.

The relatively shallow angle of the fracture would suggest that its heat treatment has been less than ideal and does differ from the manner in which earlier generation studs fail. I will dig a couple out and post photographs.

Rust is basically an Iron Oxide which comes in many forms.

Red rust is typically a hydrated Iron Oxide formed by a reaction between Iron and Oxygen in the presence of water.

Green Rust can be found on the surface of rebars in bridge decks that have suffered from Chloride attack as an alternative.

It would be correct to say that a hydrated salt could never be produced during the steel making process and would never survive subsequent processing.

Someone has clearly contaminated the surface with drops of water.

Air bubbles can occur during the steel making process and this can lead to the presence of small oxide particles within the steel ingot but this is an increasingly rare occurrence.

Normally the subsequent fabrication processes such as rolling and drawing cause small voids to weld up but small trails of oxide stringers can sometimes be found.

These oxides are often plastic at rolling temperatures and tend to extend along the length of the product as it is hot worked.

Unless the remaining stringers produce a significant reduction in the cross-sectional area of the bar/wire that will not be very damaging to its mechanical performance although clearly in the limit they are undesirable.

It is also. therefore, very unlikely that you will find defects that resemble the type of porosity found commonly in castings in any wrought product.

In the transverse direction which is what we are looking at you will tend to see very small point defects which would become much more clear if examined correctly with a Scanning Electron Microscope. (I spent far too many years locked in the dark driving one of these devices and would be happy to never see one again )

I would, in general and from looking at your photograph, tend to agree with the comments made by 'Cobalt' on the Rennlist thread.

The main initiation of the brittle failure would appear to be the dark area at 12 o'clock.

There is also a very small darker region within this defect which seems to exhibit all of the signs associated with intergranular corrosion.

I can't be 100% sure as the resolution and contrast of the image isn't great but it is typical of what could be expected.

The condition of the coating would also be relevant.

Once the corrosion crevice grows sufficiently long the crack becomes unstable and brittle facture occurs.

It is at this point the basic metallurgical condition of the base material becomes vitally important.

The 'tougher' rather than 'stronger' (although these two properties tend to be tied together) the material the greater the load it will carry and the longer the crack needs to be before failure occurs.

Once initiated by a surface defect, however, it is only a matter of time until brittle fracture will occur.

I believe that most studs actually fail during the warm up phase when the forces due to expansion become a maximum.

The provenance of the stud is also of concern and I understand that there have been poor quality 'Dilavar' studs around for a long while and as already noted the brittle fracture surface of 'your' stud doesn't show the the material isn't in the best of conditions.

The fatigue of studs has also been referred to and this is a whole other subject and as this post is already too long so suffice to say that along with con rod bolts when they are correctly torqued these studs are immune to the fatigue process.

[Steve@Rennsport]

GREAT stuff, Chris,...your input here is invaluable!!! Thank you VERY much for participating in this thread as few of us are metallurgists and you are providing one helluva education.

I'm archiving everything here.

FWIW, I've seen (early) Dilavar failures occur during cool down and it was VERY audible.

Thankfully, the latest ones (looking like all-thread) have been very durable and after 16+ years and several hundred engines, we've experienced no failures whatsoever.

[MetalDoc]

Chris, I enjoy reading your posts. Thanks for your thoughtful replys. Paul