Crankshaft material and manufacturing methods

[johnny comelately]

In amongst casting, billet, forged, vacuum arc remelted steel, there is single crystal which I think has been used in a crankshaft application.
It raises the questions of design and vibration as much as anything, but concentrating on the materials to be the most reliable, does anyone know what the current engines use?

[63l8qrrfy6]

Some grade of 3CrMoV, GKHYW or similar

[johnny comelately]

I'm trying to understand the merits of different crystalline structures as they apply to crankshaft applications.
for example, grain boundary crack initiated problems are reduced by single crystal steels.

[63l8qrrfy6]

No such thing as single crystal cranks.

[Edax]

I'm trying to understand the merits of different crystalline structures as they apply to crankshaft applications.
for example, grain boundary crack initiated problems are reduced by single crystal steels.

Yes but you get a whole host of other problems back for single crystals.

The advantage of single crystals is that they are very strong. But since the crystal planes are running through the whole workpiece you are susceptible to cleavage along a crystal plane. sC’s are therefore inherently more brittle than their grainy counterparts.

In SC you have no grains sliding against each other, but in single crystals you can get the whole material to slip over a crystal plane so they are not immune to creep.

One advantage of singly crystals is that you do not suffer from grain growth, since you only have one grain. So for high temperatures that can be an advantage. For Polycristalline materials you have to play some tricks like putting an inert material along the grain boundaries (grain pinning). Problem is that the best materials like thorium oxide are being banned.

A real advantage can be the heat conductivity, like in turbine blades. Having no grain boundaries and secondary phases really helps here.

Of course there are single crystals which have their specialist uses, sapphire for bearrings, watches barcode scanners etc (scratch resistance). Or CaF for optical windows.

But overall I seldomly come across a large mechanical application where the advantages of single crystals outweigh the problems with brittleness. Alloys where you have the freedom of controlling microstructure are usually a lot more versatile.

[johnny comelately]

I'm trying to understand the merits of different crystalline structures as they apply to crankshaft applications.
for example, grain boundary crack initiated problems are reduced by single crystal steels.

Yes but you get a whole host of other problems back for single crystals.

The advantage of single crystals is that they are very strong. But since the crystal planes are running through the whole workpiece you are susceptible to cleavage along a crystal plane. sC’s are therefore inherently more brittle than their grainy counterparts.

In SC you have no grains sliding against each other, but in single crystals you can get the whole material to slip over a crystal plane so they are not immune to creep.

One advantage of singly crystals is that you do not suffer from grain growth, since you only have one grain. So for high temperatures that can be an advantage. For Polycristalline materials you have to play some tricks like putting an inert material along the grain boundaries (grain pinning). Problem is that the best materials like thorium oxide are being banned.

A real advantage can be the heat conductivity, like in turbine blades. Having no grain boundaries and secondary phases really helps here.

Of course there are single crystals which have their specialist uses, sapphire for bearrings, watches barcode scanners etc (scratch resistance). Or CaF for optical windows.

But overall I seldomly come across a large mechanical application where the advantages of single crystals outweigh the problems with brittleness. Alloys where you have the freedom of controlling microstructure are usually a lot more versatile.

great explanation, thank you for that.
there is still a mystery to me as to crankshaft preparation particularly now on the 3 engine rule.

[63l8qrrfy6]

No mystery there - chances are that no design changes were required to make the cranks last.

If we assume that engines run on average about 3 hours per race weekend it means that last season's engine life was 15 hours while this year it has gone up to 21 hours. At an average speed of 10kRPM, the cranks had to last 9E6 fatigue cycles last year and 1.26E7 cycles this year.

In fatigue theory a component loaded within its endurance limit will last indefinitely. For steel the fatigue strength is usually defined at 1E7 cycles. It is therefore very likely that last season's cranks were designed for theoretically infinite life.

If we examine a typical woehler curve for 4140 steel (first one I could find and a common crank material) with a fatigue strength coefficient of 1745 MPa and a fatigue strength exponent of -0.07 we can calculate the maximum allowable alternating stress to be 568.85 MPa @9E6 cycles and 555.61 MPa @1.26E7 cycles.

This means that to make the crank go from 5 to 7 race weekend stress needs to be reduced by a mere 2%. For all practical purposes the cranks can be identical.

[Greg Locock]

Rule of thumb for many engineering materials: reduce stress by 10% to double the fatigue life, reduce stress by 50% to get 10 times the fatigue life.

A very smart (now retired) test engineer once told me that in his opinion steel does not have an endurance limit, the apparent endurance limit is down to bad stats and experimental technique. I wish I could remember the explanation in detail.

[63l8qrrfy6]

No engineering material really has an endurance limit, the infinite life assumption is made due to the fact that somewhere between 1E5 and 1E8 cycles the woehler curve changes its slope such that an increase in number of cycles has a negligible effect on allowable alternating stress.

It was shown that steels with a theoretical infinite life did fail at very high cycles (>1E12 or so, can't remember exactly) however such lengthy tests are impractical.

Aluminium on the other hand shows no knee on the woehler curve (slope is constant in log-log space). This is why the endurance limit for Al alloys should always specify the number of cycles it was tested at.

[Greg Locock]

random googling came up with this

http://www.abcm.org.br/anais/cobem/2009 ... 9-0507.pdf

Nowadays, many structural components are working beyond 10^7 cycles. Consequently, from the end of the eighties,
began to emerge some studies on fatigue lives greater than 10^6 cycles, and the fatigue limit in these ultra-high-cycle
fatigue range is well studied. However, only in the nineties have come to light several consistent results showing that
the steels may fail due to fatigue after reaching ten million cycles (Bathias, 1993, Kanazawa and Nishijima, 1997 and
Kanazawa and Nishijima, 1999). Bathias (1999) and Bathias et al (2001) showed results of fatigue tests on steel and
other metals and concluded that there is no infinite life under cyclic loading for these materials. According to this
author, the S-N curves obtained until 10^10 cycles did not have a typical horizontal level, that is, there was not possible to determine fatigue limit for these materials. Several authors showed the same conclusions (Marines et al, 2003, and
Bayraktar et al (2006)).

[Edax]

No mystery there - chances are that no design changes were required to make the cranks last.

If we assume that engines run on average about 3 hours per race weekend it means that last season's engine life was 15 hours while this year it has gone up to 21 hours. At an average speed of 10kRPM, the cranks had to last 9E6 fatigue cycles last year and 1.26E7 cycles this year.

In fatigue theory a component loaded within its endurance limit will last indefinitely. For steel the fatigue strength is usually defined at 1E7 cycles. It is therefore very likely that last season's cranks were designed for theoretically infinite life.

If we examine a typical woehler curve for 4140 steel (first one I could find and a common crank material) with a fatigue strength coefficient of 1745 MPa and a fatigue strength exponent of -0.07 we can calculate the maximum allowable alternating stress to be 568.85 MPa @9E6 cycles and 555.61 MPa @1.26E7 cycles.

This means that to make the crank go from 5 to 7 race weekend stress needs to be reduced by a mere 2%. For all practical purposes the cranks can be identical.

Aditionally fatigue is not something you want to play around with.

I would imagine that the main thing limiting the life of engine components is wear. As a designer I would aim for that because wear, though tricky, can be controlled and measured. There you can design more on the limit. I can well imagine that the new rules would invoke some design changes.

[gruntguru]

I would imagine the crankshaft is not problematic in the current engines. Low speed, shared crankpins, very short - no doubt the endurance limit depends mainly on how light they choose to make it.

[johnny comelately]

Honda in 2006/8 (?) had a hollow crankshaft built up using friction welding but not run due to rule changes.
one area they had cracks in (maybe the solid version) was the usual corner fillet of the main journals and it looks like adding a stress relief groove on the side face fixed it!

[63l8qrrfy6]

The stress relief feature under the pin fillet is called a Klose groove - it is not such an uncommon feature.

Cosworth have also tried a welded crank (TJ V10) however they were more ambitious than Honda and tried welding both the middle of the pin and the middle of the main journals (in an attempt to use roller bearings for both mains and big ends). Quite predictably it failed through the pin weld and they gave up.

[godlameroso]

I would imagine the crankshaft is not problematic in the current engines. Low speed, shared crankpins, very short - no doubt the endurance limit depends mainly on how light they choose to make it.

And seeing as minimum weight is specified for the crank, the focus would be on windage losses, not strength.