Discussion in 'Fiesta ST Autocross' started by Couldahadav8, Mar 15, 2014.
For your Interest and this is a good Failure Analysis case study, hard to find this kind of information and detail on the internet.
Transmission input shaft failure. Fatigue fracture.
You've described it far more eloquently than I could:
The same cannot be said for aluminum. There is no operating stress where the fatigue curve (SN curve) levels off, so eventually, aluminum will fail.
Steel alloys have a real fatigue limit where they will essentially last forever and you can look it up on a chart. The same chart will have a number for aluminum but if you design with that stress or even a fair amount lower, it will eventually fail which is very much different than lasting forever.
For aluminum, 10,000,000 cycles is an arbitrary testing length and a 20,000,000 cycle test would have a different, lower result.
For steels you probably learn everything you need to know in the first 1,000,000 cycles and prolonging the test just takes longer.
That seems like the data is being treated differently and that's what I'm getting at.
I could have sworn that the basic yield strength test was similarly different.
I've seen this failure report before. Just goes to show you what even peening the Identification on a high strength shaft can do to fatigue life. Typically fatigue cracks are created at imperfections on a surface subject to fatigue loading. Scratches, notches, etc., and vibropeen idents are typically verboten in Aerospace, military, and space applications, where I'm employed. Permanent ink would have been better. A good percentage of my time at work is spent on failure analysis.
I love failure analysis, I'd much rather strip a engine with a hole punched both sides of the block rather than a straight rebuild.
You would be familiar with United Airlines Flight 232?. The failure report and how they found the missing engine parts is a good read.
This is the graph in the Wiki article I was referring to.
In all cases, the test performed is the same. The test results are different.
I enjoy failure analyses, but being as I do this for a living, I know that a failure means lots of work, lots of customer review, and the final root cause and corrective action discussion. So, on a personal basis, I like to read them, but I know the pain involved in the process, especially when there is no assignable cause. Working on cars is why I'm in the field I'm in. I started working on cars with my dad when I was 11 and was facinated with broken parts. The why facinated me. Luckily, I was sent to a high school that specialized in metallurgy (Don Bosco Technical Institute) so I started down this road almost 45 years ago.
it's a pity that these skillsets are no longer taught to today's youth and really died out about 20 years ago in public schools... hence why today's kids cannot do much more than click a mouse. I am only 28 and I am so sad for the youth.
Lots of young kids learning hands on, just like I did in the 70s, how to become good engineers and technicians. They used to have automotive technology, but now its gone. Check out all of the other disiplines, especially materials science, engineering and technology.
I'll have to keep that in mind when I have high-school age kids - in 20 years
but yeah, the sciences are largely absent from most schools, it's nice to know something good exists relatively locally
My point is that the endurance limit for steel is where the slope goes to zero but for Aluminum they choose an arbitrary point because the slope never gets to zero. That's treating the data differently and that is why I call it a different test.
It might not seem obvious to you because real life applications don't go off the chart for a very very long time.
I work in electronics where circuits are tested billions of times in a few minutes just to check functionality, then we do accelerated life tests that take a few weeks. The way the data is treated is very important when you are trying to guarantee service life.
One of the things we worry about is electro-migration where the current exerts a lateral force on a wire which will eventually bend and break it.
By the way, I would love to know how a fatigue failure and yield failure look different, hopefully it can be seen on a scale that does not require an electron microscope.
So the official word is the Tri Point sway bar isn't actually a sway/anti rollbar, it's a chassis brace. And it is legal under 14.2.G.3; it attaches to the chassis via 2 mounting holes, braces only on one axis and is attached to the chassis no more than 2" from a factory mounting point. An actual "sway bar" would be classified/compared to a torsion bar and would be along the lines of the Honda Fit/Yaris design, or more common for our platform, the Cork Sport/Steeda/Racing Beat Torsion bar beam stiffener. I submitted another question as to the legality of running both the Tri Point "brace" and a torsion bar/sway bar; Doug is away until Monday, and will verify details and get back to me.
Hers what it looks like installed at ride height ( with the muffler clearanced)
Visually there is a obvious difference between a fatigue fracture and a yield fracture. Fatigue fractures start with a crack that grows across the surface with the load cycles till eventual sudden completion of the fracture. Yield fractures, often from sudden overload are usually brittle with some plastic deformation depending on the material and environment. For example, very hot or cold temps can have a dramatic effect on how a material fractures. Aluminium alloy turbo compressor wheel failures look like sudden brittle fractures when in fact they can be cyclic fatigue failures. A magnifying Loupe can often be enough to identify fracture types in my experience with is limited to engine failures and their components.
I am sure Westcoast can add to this with experience.
That's interesting, if I had a polished 1" round sample in an Instron and set it to the 10,000,000 cycle fatigue limit and then ran it until it breaks, is there really going to be a crack that propagates or just a creep/slight reduction in cross-section and a sudden failure?
Unfortunately I don't have an Instron and electron microscope to play with.
Maybe we are getting our wires crossed but I think I know what you are trying to get at.
I am referring to the visual difference between a fatigue fracture from cyclic load and ductile/brittle fractures from overload.
I will give you an example from my work field, engines and in this example, turbochargers.
A particular series of Cat engine I used to work required a engine rebuild at around 15,000 hours. The turbochargers would not survive that time frame and required replacement around 12,000 hours. It was referred to as a mid-life rebuild. There were other components replaced at the same time, things like coolant pumps and some other minor components.
What happen if you pushed the turbos pass 12,000 hours?, the compressor wheels would fail without exception. Referred to as wheel burst caused by cyclic centrifugal force.
The compressor wheels would crack first starting at the shaft hole and usually on the back face of the compressor wheel, bearing side. The crack would grow to a certain length and then the wheels would simple pop, always at full engine load . The fracture was classed as a cyclic fatigue failure.
It would also start with a crack, if it was a fatigue failure.
When I said illegal (at least seems to be from my take), I meant for G street (old G stock). Since you referenced 14.2.g.3, I think you are talking about the street touring rules for which the "brace" might be legal. I do not believe we can add a brace in the street class. But, I would be more than happy if I was wrong about that.
If I'm reading what you said right this implies that the Eibach bar doesn't meet the definition of sway bar either?
What Eibach bar? I ave not seen one installed yet...Heard of if "Coming soon". but as Doug pointed out, a Torsion Bar is a sway bar, and you can modify, add, or remove (1) as you like and be compliant with Section 13. He clarified that the Tri Point bar is NOT a sway bar, so would NOT be legal under section 13 but is legal for us ST guys under 14. For the STREET crowd, you need to look into the RB bar, Steeda Bar, CS bar or Eibach bar. Although I had no picture to send him for reference on the Eibach bar.
After running the ST around Road Atlanta. I highly suggest AGAINST a rear bar. If anything our cars need more rear camber. They are just too loose in the back.
Eibach anti-sway bar, Mk-7 and 7.5 Fiesta
Separate names with a comma.