F1 pankl conrod

[ringo]

The cap end and rod are two separate pieces so tensile forces cannot be transmitted through the joint, only compressive thus all tensile forces go through the bolt.
I can't see the link between the groove area and tensile force through the bolt, which is only transmitted through the bolt head and nut.

The groove may only increase compressive stress bellow the joint plane, relative to a bolt that was not grooved.

correct me if i am wrong.

I don't know whether steel bolts for racing are purposely designed not to see through to an infinite life, so it beats me if the bolt is designed to fail after x cycles. The ratio of stiffness is interesting, but i can't see the connection to fatigue in the case of 2 separate pieces unless you do a sample example.

The loads at the joint plane ,with the groove, are internal to the cap and rod and i am not sure that they relate to what the bolt sees. Those loads are only compressive.
The fluctuating load on the bolt wont be a reversed load for that reason as well, just tensile and maybe a little bending. Thus i would say the caps fluctuating loads are divorced from the bolt's.

At least this is what i visualize.

[ringo]

Call it a wild guess, but maybe the groove is for oil cooling.
The bolt doesn't have the luxury that the cap has to bathe in oil. Maybe some oil on the shank keeps down the temperature and thus the extension and radial growth of the bolt?

edit: let me finish the article you posted first!

[PlatinumZealot]

Sorry for posting twice, I tried to edit the link.

Ofcourse the net force has to be zero in the bolted joint, but you are modifying what is happening to the forces during these huge inertia loads. Do you really think you could clamp up a bolt to say 80% of its yield then ask it to deal with another 40000N every other revolution so say 150 times a second. It wouldn't stand a chance.

http://www.kxinc.com/content/Kx%20Broch ... _Brief.pdf

That looks like a brief paper on some bolted joints and I see it has a couple of formulas there about fluctuating loads in joints, not derived though.

I am reading the posted link.. It speaks about "joint lift-off" of a Bolted joint. The Bolted joints described in the paper are bolted flanges that carry eccentric loads. The eccentric loads tend to cause the flange itself to act like a pry-bar acts to pries the bolt off, causing bending which not only separates the joint but is an additional load on the bolt. - I understand that part.

It then goes on to say that stiffening the flange is not the solution to the problem as stiffening the flange may only serve to make the prying action more effective. So, that:

1. a good balance in joint stiffness vs bolt stiffness is the way to go.

2. It also says that the line of action of the force should be brought closer to the bolt centre lines to reduce the bending action. makes sense - all good.

Some issues though:

1. This paper is focusing on saving the bolt during joint separation. hence keeping things intact.

a) it says solution 1 reduce the line of action of the force. - You are greatly restricting from doing this in a connecting rod application. The body of the cap literally wraps around the bolt. can't change much.

b) It says reduce the stiffness of a BENDING flange. Think of like a crow bar. It's harder to pries a lock with a soft crow bar. Sounds like it can work but one has to think carefully about this. This is the flange bending away instead of the bolt.

This is a situation that doesn't necessarily deal with the shape of the bore. But, I can see it working to keep the connecting rod in one piece, that is if it is actually the situation in a connecting rod setup. So what we have to do know is verify if the flange of the connecting rod bends the bolt at peak load AND then we add the notch and see if the flange bends away from the bolt at peak load.


If true (flange beding and joint separation in a connecting rod) that means that the notch is to increase the life of the bolt. Not the life of the connecting rod nor to help the oil film.

Again, that is if the flange on the connecting rod separates and bends the bolt... I doubt it will separate but maybe there is some bending shock waves transferred to the bolt. It is a nice detour so let us see...

Somebody whip out the CAD software!

[ringo]

After reading the article, it basically goes back to what riff raff mentioned.
And what smikle has posted; until the pre load is exceed there isn't much concern for the cap stiffness. Under compressive load, ie if the external load is bellow the preload, which it should be, a stiffer flange reduces the load carried by the bolt.

Fb = P + kb/(kb + kj)*Fe

Fj = P - kj/(kb + kj)*Fe

where
Fb = Load carried by bolt (lbs)
Fj = Load carried by joint (lbs)
Fe = External load (lbs)
P= Preload (lbs)
kb = bolt stiffness (lbs/in)
kj = Joint stiffness (lbs/in)

Notice, from equation 1.0, when kj >> kb the bolt loading from external loads is virtually zero as long as the joint remains in compression. For this stiffness condition, however, the joint will come out of compression when the external load reaches the preload value (e.g., reference equation 2.0 and set Fj = 0.0,). For conditions where kj > kb the bolt will carry a portion of the external load, but it also requires an external load greater than the preload for the joint to come out of compression.

Whether the pre load is exceeded, i have no evidence, but i suppose it shouldn't.

[xpensive]

Just some design-engineering basics:

- The purpose of the bolt-preload is to be higher that the anticipated xternal load on the joint. That way the stress in the bolt itself will never change with xternal load, unless said preload is xceeded, only the contact force between the mating surfaces changes with xternal load.

- The "groove" is oviously there in order to determine two contact patches, one on either side of the bolt, in order to minimize the effect of a less than perfectly flat and parallell mating surfaces, a very common method in all sorts of bolt-joints.

[PlatinumZealot]

That is interesting, the contact patch part.

Any pictures?

[xpensive]

Like when bolting a bearing housing;

http://www.skf.com/skf/productcatalogue ... =1&lang=en

Perhaps not the closest of xamples, but it gets you an idea of the principle.

[Richard]

The post is about the design of the Pankl con-rod shown.

I got that bit. Is it currently being used?

[riff_raff]

The point made about relative stiffnesses of the bolt and clamped conrod sections would seem to be important. As noted, the preload strains in the bolt and conrod sections are likely not equal, unless the materials and section properties are identical.

Making the conrod sections less stiff than the bolt would be of benefit since such a joint arrangement would be less likely to separate at the beam/cap face under load reversal. Consider it like this, if the bolt was very stiff axially, and the clamped conrod section were much less stiff in compression, then the assembly preload would produce a greater strain in the conrod sections than the bolt. And when the bolt gets loaded in tension under dynamic conditions, the reverse is true. The stiffer bolt will deflect less than the softer conrod sections would naturally want to relax for the same force. Thus the conrod joint faces would more likely remain in contact.

Here's some numbers, just as an example:

case 1- stiff bolt, soft conrod sections
Bolt preload axial strain= .01mm
conrod clamped section axial preload strain= -.02mm
bolt axial deflection due to dynamic loads= .01mm
result= bolt deflection under dynamic loads does not exceed conrod section preload strain and joint face does not separate.

case 2- soft bolt, stiff conrod sections
Bolt preload axial strain= .02mm
conrod clamped section axial preload strain= -.01mm
bolt axial deflection due to dynamic loads= .02mm
result= bolt deflection under dynamic loads exceeds conrod section preload strain and joint face separates.

It's a bit simplistic, but I'm sure you get the point. I ignored structural spring rates, but if you wish to consider them, they are linear for these cases and are force per unit change in length (ie. lbf/inch). So in short, if that groove makes the clamped conrod sections less stiff in compression, it would be of benefit.

riff_raff

[F1_eng]

xpensive,

"The purpose of the bolt-preload is to be higher that the anticipated xternal load on the joint. That way the stress in the bolt itself will never change with xternal load"

This is not true. These graphs help to visualize the forces and strains in a bolted joint.
http://www.boltscience.com/images/jdia4.gif

Some of you might be starting to see it now. That paper wasn't meant to be an explanation of exactly what was going on in a con-rod joint, it had a couple of decent stiffness formulas, but it seems it might have helped you to understand bolted joints a bit.
And no, the bolt pre-load would never be obercome by external load. The bolt is only protected from some of the external force if the cap has some strain left to be reduced. Otherwise, the bolt is on its own in dealing with the total force.
Plus you would have other major issues if the joint starts to separate.

[Formula None]

And no, the bolt pre-load would never be obercome by external load. The bolt is only protected from some of the external force if the cap has some strain left to be reduced. Otherwise, the bolt is on its own in dealing with the total force.
Plus you would have other major issues if the joint starts to separate.

Making the conrod sections less stiff than the bolt would be of benefit since such a joint arrangement would be less likely to separate at the beam/cap face under load reversal. Consider it like this, if the bolt was very stiff axially, and the clamped conrod section were much less stiff in compression, then the assembly preload would produce a greater strain in the conrod sections than the bolt. And when the bolt gets loaded in tension under dynamic conditions, the reverse is true. The stiffer bolt will deflect less than the softer conrod sections would naturally want to relax for the same force. Thus the conrod joint faces would more likely remain in contact.

Good posts. Thanks again for the explanation.

F1 eng, what's your background?

[xxChrisxx]

And no, the bolt pre-load would never be obercome by external load. The bolt is only protected from some of the external force if the cap has some strain left to be reduced. Otherwise, the bolt is on its own in dealing with the total force.
Plus you would have other major issues if the joint starts to separate.

I've not read everything (don't really have time with a half hour dinner), but were you referrign to joint stiffness with regards to the groove?

Out of interest how do you specify your preload? By torque or bolt extension?

In my old job we were specifying bolts on a subsea connection, it turned out be be a horrendous pain because we specified a torque but a different lubrication was used by the customer. Caused all sorts of haedaches when we didn't have enough preload.

[PlatinumZealot]

xpensive,

"The purpose of the bolt-preload is to be higher that the anticipated xternal load on the joint. That way the stress in the bolt itself will never change with xternal load"

This is not true. These graphs help to visualize the forces and strains in a bolted joint.
http://www.boltscience.com/images/jdia4.gif

Some of you might be starting to see it now. That paper wasn't meant to be an explanation of exactly what was going on in a con-rod joint, it had a couple of decent stiffness formulas, but it seems it might have helped you to understand bolted joints a bit.
And no, the bolt pre-load would never be obercome by external load. The bolt is only protected from some of the external force if the cap has some strain left to be reduced. Otherwise, the bolt is on its own in dealing with the total force.
Plus you would have other major issues if the joint starts to separate.

You are the one who is starting to see it. That graph and what riff-raff said contradicted what you said before.

And lets not get too blinkered here everyone. We all know you have to pre-load the bolt and have a certain flange stiffness for a good quality joint. That's done and dusted.

The paper you F1eng posted is about a flange bending in bolted joint before and during separation. What I wanted to do is see this is the case in connecting rod. We know separation is not what happens (or else you got your self a busted engine block!). So we have to investigate how the square channel in the flange affects the bending load on the bolt.

[PlatinumZealot]

Maybe we can do a solid-works or (ANSYS) analysis ourselves?

[ringo]

The point made about relative stiffnesses of the bolt and clamped conrod sections would seem to be important. As noted, the preload strains in the bolt and conrod sections are likely not equal, unless the materials and section properties are identical.

Making the conrod sections less stiff than the bolt would be of benefit since such a joint arrangement would be less likely to separate at the beam/cap face under load reversal. Consider it like this, if the bolt was very stiff axially, and the clamped conrod section were much less stiff in compression, then the assembly preload would produce a greater strain in the conrod sections than the bolt. And when the bolt gets loaded in tension under dynamic conditions, the reverse is true. The stiffer bolt will deflect less than the softer conrod sections would naturally want to relax for the same force. Thus the conrod joint faces would more likely remain in contact.

Here's some numbers, just as an example:

case 1- stiff bolt, soft conrod sections
Bolt preload axial strain= .01mm
conrod clamped section axial preload strain= -.02mm
bolt axial deflection due to dynamic loads= .01mm
result= bolt deflection under dynamic loads does not exceed conrod section preload strain and joint face does not separate.

case 2- soft bolt, stiff conrod sections
Bolt preload axial strain= .02mm
conrod clamped section axial preload strain= -.01mm
bolt axial deflection due to dynamic loads= .02mm
result= bolt deflection under dynamic loads exceeds conrod section preload strain and joint face separates.

It's a bit simplistic, but I'm sure you get the point. I ignored structural spring rates, but if you wish to consider them, they are linear for these cases and are force per unit change in length (ie. lbf/inch). So in short, if that groove makes the clamped conrod sections less stiff in compression, it would be of benefit.

riff_raff

Sound reasoning, though i think if the bolt extends more the the cap and rod for a given load, the cap and rod will naturally regain contact by relaxing into the gap to the bolt head and nut.
Only when the cap and rod reach their unloaded states will they separate; which is an exceedance of the preload.

The time it takes for the cap to relax into faster extending could be a consideration.