As with all mechanical components, there is a finite tensil or structural durability component to it.

For Example, a Stock Camaro w' LT1 has approximately 300 HP. The stock clutch, transmission, driveshaft, u-joints, differential, and axle components are all capable of sustaining occasional burts of maximum power. In addition, the "over-engineering" of most of the components allows for substantial power increases without failure.

However, at some basic level every component will be stressed and fail. That is why we have 4-bolt mains, 9-inch rear ends, ARP studs, etc.

What are the most common components to fail at given HP/Torque ratings:

1) 300-350 HP
2) 350-400 HP
3) 400-500 HP
4) 500+ HP

It seems to me that playing HotRod is a never ceasing litany of breaking one component after another in the driveline succession after each subsequent power increase.

 

Sheer power isnt gonna break parts all that often. driveline failures are due to grabby clutches, and sticky tires. the shock load is what breaks them.

 

You can drive around all day long on a 1,000 HP setup with a stock F-body drive train.

However, once you start doing 6k clutch dumps, 1.4 second 60 ft times etc etc - your weak spots will surface in a hurry

It all depends on what your going to be doing with the car really.

 

uhhhhh hmmmmm dpmt be foregetting the OPTI im on my THRID one this month mother f'n pieces of.... i cant believe my car has gotten to the point were it can shatter an opti in less that 200miles those things are the biggest piles of .... and if i could get this damn LTCC to work maybe everything would be ok

 

I would not include the Opti in a list of components that is HP limited. Like any distributor/coil ignition, it is going to be RPM challenged, and may not develop adequate voltage with the stock ignition. But it could care less what it sees in the way of HP, particularly after you get the high voltage stuff out of the case. The quickest LT1 ran a stock vented Opti at 1,125flywheelHP, with no problems. That was a "optical sensor only" application, similar in concept to the LTCC. If you lose an Opti in 200 miles, there was a defect in the part (did you check the rotor hold down screws - they have been known to come loose on brand new units), or it was installed incorrectly.

 

Agreed. Todd Carpenter runs in the Unlimited Class when we race the Silver State. He's running over 210 MPH for 90 miles... he's running a 10-bolt and rebuilt tranny.

But in the Silver State, you're going for top speeds not drag racing. It's taking off the line easy... what is your intent with the HP?

 

I know there are some engineers here, but none of them have jumped in for an "advanced" look at this. So I will at least try to get them interested. There are number of different properties of materials that can be defined which relate to the "strength" of a component. These include properties such as Young's modulus of elasticity, yield strength, Poisson's ratio, elastic and plastic deformation, tensile strength, compressive strength, and strain hardening behavior among others. All contribute to the behavior of a component when it is stressed.

When a sudden, high load is applied the material will first deform and then may completely fail if the load is high enough. If the stress is less than the yield strength, the component will deform under the stress but return to it's original form. If the yield strength is exceeded, permanent plastic deformation occurs. A stress which causes plastic deformation will permananely weaken the component and lower the yield stength dramatically. Stresses of a lesser magnitude cause fatigue by other mechanisms. Loads less than the amount needed to cause outright failure contribute to fatigue. Over time, repeated stresses that aren't high enough to cause failure when applied once will cause failure as the material fatigues.

The parts we are talking about here usually do not fail the first time they are stressed or even after multiple stresses. IOW, the force applied is much less than the yield strength. But over time, the part will fail. The larger the stress, the fewer number of cycles the component will tolerate without failure. So it doesn't make sense to talk about a hp limit without somehow incorporating the idea of repeated stress. This concept is why certain parts should be periodically replaced on a highly stressed setup. Valve srpings are an obvious example. Aluminum rockers are another (aluminum usually has less fatigue resistance than a steel part of similar tensile strength).

The T56 is a good example. I never heard of a new one breaking in an F-body. It takes repeated stress over time to cause failure at the power levels we are talking about. Ditto for stock steel drive shafts, though I have seen aluminum shafts fail the first time out. Same with cranks, rods, axles, etc. Rear end gears as well. Hopefully, some of the engineer types will help clarify. But what I am trying to say is that you can't meaningfully specify a hp level at which a component is good for without somehow accounting for how many fatigue cycles you plan to subject the part to. Practical experience does give something to go by though. The weakest links appear to be the A4 tranny, the rear end, and the aluminum driveshaft. I have also seen axle failures. But how long they will last at a given power level is really what one would want to know, and the data is not out there to provide this information.

Rich Krause

 

Rich, you said it all eloquently. Read Rich's entire post a couple of times guys.


One thing you described was "endurance limit" or a stress level where the part can be cycled almost forever and it won't fail. That's well below the yield point, and where OEM parts are designed.

As you increase the loads (or stress) the weakest link fails first, and if you strengthen it, the next weakest link, etc.

An engineer once told me that "Metal is like a woman; it remembers exactly how often and exactly how hard it was stressed, and when it reaches it's failure point it often fails catastrophically with no outward indications." Although not quite PC, it's a fairly good analogy.

Driveline parts are in the same category as engine parts as far as stress and strain (deformation) are concerned. Different racing engines are designed for different numbers of life cycles:

A Winston Cup engine that ran the whole 500 miles at Pocono yesterday (and in qualifying and Happy Hour) turned over about 1,200,000 to maybe 1,500,000 times in anger. Gibbs' engines didn't have much over 1,000,000 cycles in them, evidently.

Contrast that with a Top Fuel engine that runs about 600 revs in anger then gets rebuilt. Of course the TF engine puts out about 5-1/2 times as much hp per cube as the Cup engine. The metals used are of approximately the same strengths.

In a 10 second 5000-8500 rpm drag engine, 1.5 million cycles is about 1100 runs. Bet few go that far between rebuilds.

My $.02

 

How does the yield strength of a carbon fiber drive shaft compare to aluminum or steel?

 

It depends on the composition of each.

Some CF composites are around 40-45,000 psi Y.S.
Aluminum can vary from under 20,000 to above 50,000 psi
Carbon steel (1026 DOM tubing) can be 50-75,000 psi. 4130 (chromemoly) can be much higher.

It's not just yield strength, but size and configuration that matters, among other things.

A long, small diameter tube tends to whip when spun fast. There is a Length divided by Diameter (aka "L over D" or L/D) ratio than isn't usually exceeded. For example use an L/D of 15. A 45 inch shaft could then use 3 inch diameter tubing, but a 60 inch shaft would need 4 inch tubing. The weight of .065 wall steel tubing increases from 2.03 lb/ft (3 inch) to 2.73 lb/ft (4 inch), so driveshaft weight and rotatiing inertia increase quickly.

Aluminum weighs about 1/3 that of steel, and the alloys commonly used for tubing are about 1/2 to 1/3 the Y.S. of steel, so the 4 inch alum. tube probably has a .187 (plus) wall thickness.
If the U-joint end is also aluminum it has to be larger than steel to be as strong. So, for equivalent strength, the Alum. shaft isn't all that much lighter. GM uses alum, especially in long wheel base pickup trucks so they can use 4 or 5 inch diameter 1-piece driveshafts and not need 2-piece ones. Large diameter, thin wall aluminum shafts don't clear the floorpan in a car (F-body).

CF itself realy doesn't have a yield strength: it doesn't yield, it just breaks when it reaches its tensile strength. When it is made into a part, epoxy is used to hold the CF in place, so the resulting composite has some Y.S. Joining u-joints to the CF composite tube is usually the biggest challenge.

Metal Matrix Composites (MMC) are starting to be used for driveshafts. With these you can get strength, weight , and relatively low costs (at OEM levels).

 

Enter the world of S-N curves. Some metals have what is called fatigue limits and others do not.

Some metals can be cycled under low stress levels for years on end, and never have a fialure. Such is the case with suspension springs. Springs go through constant loading and unloading; and constant deformation. So in theory, if you choose the right material for the right loads, you'll never have a failure. But that's only in the fine-and-dandy world of simulations.

In the real world, what you have to watch out for is dynamic loading. Many materials will respond differently to equal stresses according to how fast those stresses are applied to the material.

 

OK, you just hit on one of my pet peeves, so maybe you can explain why it does or does not make sense.

Yes, the YS of 1018, 1020… will be roughly in that range, but a similar part made of 4130 will have roughly the same YS (well, usually 3-4000psi better, which really isn’t that much stronger) assuming that it is not hardened. Add the fact that most of what is sold as ‘high performance’ chromemoly parts is not hardened and not even normalized after welding you have a part that isn’t any stronger then a comparable mild steel part with brittle areas near it’s welds.

The part of this that really bugs me is that many thin walled 4130 suspension parts and thinner walled, smaller diameter 4130 roll cages are considered ‘safer’ then similar mild steel parts.

Even if someone was going to go to the effort of hardening something like a 4130 DS you’d have to keep the wall thickness down (4130 doesn’t really take more then a shallow heat treatment).

 

You make a very good point, Mark.

I picked Drawn Over Mandrel (DOM) tubing for a reason; it's stronger than plain welded carbon steel tubing and it has very uniform wall thickness, so it makes a good drive shaft.

Welded tubing, such as water pipe or exhaust tubing is made from a flat strip of steel which is rolled into a tube much like you would roll a cigarette but the ends of the strip are butted together and heated until they fuse together (welded). The weld flash is scraped (scarfed) off the outside, and sometimes the inside. The weld is at least as strong as the parent metal, so strength of welded tubing isn't an issue.

DOM starts as welded tubing, but then it is pulled or "drawn" over a mandrel which expands its size a few %, but more importantly, work hardens the material which increases its strength. The same thing applies to cold finishing steel bars or strips. Steel bar is first shaped when it is (red) hot and doesn't have much grain flow. It is called "hot rolled" at that point. If it is then pulled through a die which reduces it's diameter, the grains distort or "flow" along the length of the bar and is now called "cold drawn" or "cold finished" bar. It gets somewhat harder, and it's tensile strength and yield strength increase. Interestingly it also machines easier and faster than hot rolled steel.

DOM is usually made from 1016 to 1026 steel, while plain welded tubing may be 1010-1016. The last 2 digits indicate how much carbon (.26% in 1026) is alloyed in the steel. Carbon content (up to a point) increases the strength. 4130 has .30% carbon, 4140 has .40%, etc, and are known a medium carbon steels. The 41 indicates chrome and molybdenum.

Because DOM is about .20% ("20 points") of carbon it approaches the strength of medium carbon (30-50 points) steel.

4130/4140 is actually quite deep hardening, due somewhat to the cr and mo in the alloy.

4130 tubing is "seamless". It starts life as a bar which is heated and a hole is poked thru it. Really. It's then drawn, annealed (softened with heat and slow cooling) and redrawn to get the finished size and strength. While the wall thickness is usually fairly consistant in the thinner wall sizes, DOM is usually better, and makes very good shafts.

If 4130 is welded properly and normalized, another relatively low heat treating process, it retains it's strength. One of the reasons many older (and homebuilt) aircraft used 4130 space frames was because when gas welded properly it doesn't lose strength. This can be done by almost anyone and doesn't need shielding gas. Today TIG is the preferred method, but oxy-acetelene (gas) welding still works.

I too am suspicious of 4130/4140 fabricated parts that have been welded unless the heat treating is certified and tested.

IMO, sanctioning bodies have a difficult time telling DOM from welded tubing in a finished cage. The strength differences can be up to 40-50%. They are able to test for 41xx alloy, so specifying that on things like Top Fuel is smart. I have heard that NASCAR will go into a Winston Cup fab shop, and take a piece of tubing right from the cage builder's hand and test it to determine its strength. I believe they specify DOM in the rules.

My too-long $.02

 

Thanks for the science lesson Jon
Good stuff

 

I agree- VERY kick *** reading.
I just got back from a lecture with a GM NASCAR engineer who designs the blocks. A lot of this relates to what he just said.

How crankshafts are designed to last the legth of the race and thats it. if they last longer- they are too heavy. It is a compliment to the engineer if they last the whole race and are never used again. Kind of stomps the whole "It will break or it wont break" way of thinking huh? Perhaps not if but when is the approprate questions in many situations.

Also another thing Id like to ask Mr. Baur. Is there any such thing is a "strong" metal? Isnt saying "its strong metal" like saying "its a good car?" Although my mother is convinced her maxima is a good car, I beg to differ (WHERES THE REAR DIFF!)

Back to what I was asking- "Titantium is stronger than XXXXXXX metal" Well then, why dont you run a titanium cam and let me know how that turns out In recent debates i've overheard (and some i wish i didn't) I always hear "is that metal strong enough"

Brittle, maleable, hard, soft, dense, porous and all the charactoristcs Rich mentioned in his first post tend to be adequate- but (and perhaps this is my question) "what is a strong metal?"