Simulating large force transmitted via hydrodynamic bearing in static simulation

[peterblais]

Hi guys,

I'm working on a part - the big end of an automotive connecting rod - to be specific. The issue that I'm having is that currently, I am simulating with the rod, it's cap, two dowel pin sleeves for alignment, and a pair of bolts. I'm using no penetration contact between the rod and the cap, which allows the bolted joint interface to open slightly as it should. I'm applying my principle loads to the big end cap via a solidworks "bearing load" with a sinusoidal distribution.

The problem I'm having is that the results seem most likely incorrect to me. The deformation at the big end is more then the oil film between the components in operation.

Now, my assumption here is that the crank pin minimizes the deformation which is physically possible. When the "sides" (near the bolts) start to squeeze in on it because the part is trying to elongate- the oil film picks up some of that loading and forces a more round condition on the part.

My concern is that I am over-estimating stress levels in this portion of the component. We have not had any failures, but over the years I have been refining my simulations to whittle away at mass...

I tried modeling a dummy crank pin and setting it in there as a body floating with .001" clearance (total guess there, oil film is really more like .001 to .003" depending on the engine builder / how worn out the crank is the customer is using)... Solver did not like that... Anybody have any suggestions here? Maybe a thin dummy component with a young's modulus that would simulate that pad of oil in there, backed up by a solid steel crankshaft pin- just fix that in place and apply a standard bearing load to the opposite end which is not being studied...

 

[MikeHalloran]

Just for fun, I'd try to model a one-piece big end.
Leave the holes and bolts in place, just pretend it's a 'cracked' big end that was never cracked. See how that deforms before worrying about the cap/rod joint.

Can you model the oil film as a sort of annular capsule that's inflated with a constant pressure? Sort of like a hydrostatic bearing? Then maybe divide it into quarters and adjust the individual pressures to reflect whatever empirical data you can find?

Just random thoughts; I miss having access to SW.



Mike Halloran
Pembroke Pines, FL, USA

 

[peterblais]

Mike,

Thanks for the suggestions. Even with it "one piece" the part still turns into an egg without the crank pin in there. My conclusion is that the big ends start to deform, the oil film gets pinched down near the cap / rod joint, and starts supporting the rod in those areas. What I ended up doing was mating up a crank pin with the oil clearance (.0015") in the big end, using some spring supports to "hold" it in place. I used a no interference contact set between the components- once the rod starts deforming, it collapses down on the crank pin and this limits the shape to what is possible.

I doubt the actual deformation is this high since hopefully the oil film would not completely collapse and go to zero- but I'm not about to start guessing at it's actual thickness. Assuming it goes all the way to zero gives me a bit of extra safety factor. The amount of deformation you allow at the cap-joint greatly effects stresses from bending in the middle of the cap, and around the bolt hole pockets.

With that sorted, solving it as one piece, versus 2 pieces + rod bolts, gives only a relatively small change in stress. The two piece version shows some more deformation because of the joint opening up slightly on the innermost edges and IIRC (this was a while ago) increased stresses ~5% or so. Calculating with the bolts in though is nice, you can get a bending stress for your rod bolts out that way.

It's good enough I can sleep well.

 

[Tmoose]

" which allows the bolted joint interface to open slightly as it should "

I think if that means the bolted joint faces separate then the rod bolts will fatigue within a few months, if bearing mayhem does not bring things to a halt in the first few minutes.

============

Can you post a picture of the con rod? I'm wondering if the geometry is unusual.

Hydrodynamic oil film thickness is highly load dependent. It gets mighty thin if the load is high, like in an engine. No "centering" worth mentioning in my opinion, despite what the late great Don Bently said.

http://www.astbearings.com/assets/images/Bushing-Hydrodynamic-Lubrication.jpg