Originally Posted By: Shannow
Originally Posted By: lomez
any part that sees constant hydrostatic or hydrodynamic oiling..... last forever? I mean, the metal is by definition not touching. Barring some type of cavitation, in theory, the parts/gears/bearings, etc should never wear out. Yet.....they do. ??
The same would seem to be the case with modern oils and boundary protection. In theory, shouldn't most engine never wear out (exceptions perhaps for a chain....but even that's ify)?? How does a ring and pinion gear wear out....assuming constant oiling? Many would say "at start-up" (as in before the oil flows to the parts in question)....but still, doesn't boundary provide protection? I can see cylinder bores, as they don't really get thorough oiling...but that's about it. How does a bearing EVER wear out?
Just wondering how, assuming complete oiling.....parts ever fail. Looking for some education on the subject.
I've pulled apart turbines with 150,000 operating hours, and pristine bearings, virtually no wear, and I've pulled apart turbines 15,000 hours after an overhaul, and every bearing has been wiped (still running)...for no apparent reason.
I'll use the turbine example, as they can (I hope) be visualised as to some of the issues that can arise....consider 200 tonnes of shaft rotating on ISO 32 oil, 6-7cst at operating temperature (who said I have a thick oil bias...hmmm???)
Take 5 shafts, weighing from 25 to 60 tonnes each, each shaft having two bearings (1 each end), bolted together to make a single long shaft of 70-80 feet, supported in 11-14 bearings, on a film of 7cst oil at 3,000 RPM.
http://dc354.4shared.com/doc/HPt40m2h/preview.html gives an idea (although the picture shown, they have "cheated" by having only one bearing per coupling)
The bearings range from maybe 12" to 21" in diameter, and 8-20" long, and typically have a radial clearance of 0.00075" per inch of bearing diameter...scaled, those dimensions aren't that different to IC stuff...
The bearings are hydrostatically lifted before turbine roll, and at around 300RPM, the hydrodynamic oil wedge provides enough lift for the shaft to ride on. Oil pressure from the hydrodynamic lubrication is about 1,000kPa (145psi), with the hydrostatic lift 2900psi to get things rolling. hydrostatics is applied on rundown from 400RPM down to stop (actually 4RPM for 4 days after a rundown), but the bearing can safely run from speed to stop without damage, can't be brought back up to speed without hydrostatics without doing damage.
IC engine parallel, is that you have no hydrostatic lift, so on start, the shaft has to climb the bearing, then slide down the face to initially establish a hydrodynamic wedge, and will complete a number of revolutions to do so...it's a start thing, and once running, you hope that the designer kept things hydrodynamic, and apart
Look at the whole train of turbine from on high vertically...all the bearings have to be in a line...if they are not, the shaft will find it's own centreline, which will be an average of between all of the bearings, depending on how they are laid.
Given that there's only 0.004 to 0.008" between the shaft and the bearing sides, and it's hard to get the shafts closer than 0.002" centre to centre, all of the side clearance can be sucked out of one or more bearings with a not good alignment, causing the hydrodynamic wedge to be moved from it's normal position to some other "o'clock" position, and when running, be trying to centralise the shafts amongst them all (bends the shaft slightly)...can see this, and inadequate shaft to shaft centreline alignment by watching the hydrostatic pressure gauges while the shaft is turning slowly. One bearing will have double the pressure of it's mate, then they swap.
IC engine parallel is crankshaft stiffness, and bearing alignment.
If the tunnels aren't perfect, then the designed in loads on the bearings could be in the wrong direction entirely, or possibly doubled in magnitude...long life can't be assured under those conditions, but failure isn'tlikely in seconds
Turbine train looked at side on, imagine picking the shaft up like a chain from it's ends. It will "sag" (like a power line between centres), in a shape that's called a "catenary". That is the natural form that it will sit in.
So the bearing heights (another alignment) have to be different, so that the combined shaft sits low in the centre, and progressively higher at each end...that way there is no low end/high end, and when spinning, it wants to flop into the middle, and not put an end thrust onto the axial thrust bearing.
Mess the heights up, and like the sideways alignment, different bearings are loaded/unloaded, which can cause problems.
Consider also that the bearings have to be set up all at funny angles now to follow the curve of the shaft, or the load will be uneven, and the bearing will wipe on one side, or not jack (lift) under hydrostatic lube conditions...
IC engine parallel is alignment again...concentricity and parallelism...turbine bearings can deform under such misalignment and make themselves work for a while...IC trimetals not so much.
Alignment can upset thrusts, we use a little steam pressure imbalance to load the thrust...engines (like the SBC) can misalign the crank and cam bores so that the chain can slightly preload the thrust
Unbalance on rotating components will mean that the shaft is not spinning dead centre on the oilwedge.
1.5lb imbalance on a 35 ton rotor spaced 1.5' off the shaft centreline will give approximately 100um vibration, about the upper limit for long term steady (I've run double that for a year when necessary).
100um is 0.004" peak to peak, which means that the oil wedge is being "pumped" by 0.004" every revolution, and much of the side clearance goes as well, the oil wedge having to walk around the bearing to keep everything central.
White metal can fatigue from this.
Bearing white metal to shell bond can fail.
IC Engine analogy, same with balance, it uses up clearance when cranking and at low speeds. The oil wedge will try to centralise it, but the "pounding" will fatigue bearing surfaces
IC engines have major cyclic loads to deal with as well as balance, lubrication only during a partial arc of their travel, so can be worse.
IC engines have gyroscopic things to deal with as well - imagine the gyroscopic loads due to the flywheel when doing a stationary spin
Debris can get in, and is often more than the oil wedge thickness. It will feed into the bearing, and get trapped. White metal is designed to allow it to embed in the bearing metal, and hopefully get away from scoring the steel journal
IC engine example...same idea, smaller clearances...better filters, however
Cavitation - oil is dragged into the clearance gap, and as there is diminishing space increases in pressure, and some leaks out the side. As the oil passes through the wedge, the gap increases, reducing pressure, often to severely below ambient. Extreme conditions, can create cavitation bubbles, and chew up bearings.
IC engines same
Microdieseling - aerated oil as it is compressed into the wedge can increase in pressure rapidly, causing microdieseling (oxidation products then are sludge precursers, and use up anti-oxidants)
IC engines same
Sparking - steam (an insulator) passing through the turbine blades creates many thousands of volts of static, which can discharge through the bearings. To this end, there's an earthing brush to make a single low resistance path.
Failure can mean bearing erosion, and oil sludge
IC engines - can have electrical discharge in some synthetic media, no reall parallel
Instability - sometimes a combination of bearing load, oil temperature, and cyclic vibrations
IC engines - you will never ever know, but that could be why you spun a bearing
Like I said, I've seen them come apart after 150,000 hours service, and look brand new. Seen them come apart with whole teaspoons of metal wiped off the bottom, then dragged around the bearing and deposited on the other side where it won't fit into the clearance - still ran. Done an alignment to repair a single bearing using a crane weight to estimate the alignment pressure to save a week of disassembly.
Perfect world, they will all last forever.
World is never perfect, and when you buy an engine already buttonned up, you will never know what isn't perfect, until it either tells you through failure, or you wear something else out.
Bravo sir!
What a wonderfully complete explanation.
