rotarygod

This may sound like a weird question so hear me out. What is a bearing in an engine anyways? We don't use ball bearings or roller bearings. I admit there are some very small uses for these in specialty engine applications but for the most part in an engine we don't have them so we'll just leave it at that and go from there assuming we don't have those. The bearings in an engine are really nothing more than steel cylindrical inserts that are pressed into the rotors and stationary gears. These sleeves have a groove in them (some bearings don't) for the oil film to travel in. A bearing never touches metal on metal which is why bearings don't ever break in. They just need to be heat cycled to be broken in. The thin film of oil is what keeps them apart and what keeps everything happy. From time to time a bearing may come lose and "spin". This causes all sorts of issues but on a rotary that ususally means you are replacing the component that this happened to as well as the eccentric shaft plus whatever else it decided to take out with it. The eccentric shaft doesn't have a bearing installed on it. We just use the shaft itself. Same thing here. Let's do it on the other side too.



Why do we need these inserts? Why can't we just machine a tighter tolerance into the rotors and the stationary gears so they work as the bearing themselves? Remember it's really the oil that does the work and not the bearing. Nothing should be touching. This would really simplify things. When you press a bearing in, it can distort in shape a little bit and get out of round slightly. That's bad news. Hopefully if this happens the engine builder will make the necessary adjustments. We can easily avoid this if we just don't have one. If the stationary gear and rotor are machined good, we shouldn't ever have an out of round, out of true issue. If something fails we still have to replace everything anyways so longevity isn't an issue. If anything this should be more reliable because we can't spin a bearing. This is simpler, has less room for error, still does the same job, and has no drawbacks that I can think of. Why couldn't this be done? The reason I bring this up is because this is actually how NASCAR engines work. No "bearings". Yes they have very pricey engines but that shouldn't really matter here. It's the proof that the idea works. Machining the tolerances into everything is no more difficult than what they do now. They still have tolerances they machine to on the stationary gears when they are machined now. They are just larger to allow a metal sleeve bearing to be installed. It would be so much easier to just adjust this machining spec to get rid of the bearing.



It all makes sense to me. Admittedly I can't think of a way to make this work unless Mazda themselves do it. What do you think? Why not?

ColinRX7

I don't see why you couldn't machine each part to be within tolerance, and eliminate the bearing, but is it worth all the effort? You would get the same effect. For what it takes to make it happen, you could have paid for another rebuild with a spun bearing. But how often do rotary engines spin rotor or stationary gear bearings?



Also what about bearing material? I don't see any reason why it wouldn't work, but having a bearing gives you an opportunity to use a more ideal material without having the permanent piece made of the same (costly?) material..







Bearings have been around for years, I don't see my mazda would stray from that.. Less worries, another part to replace which could potentially save a "permanent" piece if the bearing didn't exist, less permanent parts to have scrapped...



The way I see it, is part for mass production, and ease of mass production assembly, less in-plant risks of damaging a permanent surface (what's a bearing cost to scrap if it got damaged as opposed to tossing an whole cast rotor). Keep in mind though I'm talking about the companies that are machining the parts for mazda, reducing in-plant error costs, etc.



If you unfamiliar with mass production business, for example, if you have a holset turbocharger on your cummins turbo diesel in your brand new dodge pickup, the hotside was cast and machined in Guelph, Ontario, Canada, by Linamar corp. Cummins will approach a business, and the business will say "we can make you this many for this price". It's those businesses that are very particular about reducing in plant costs, because it basically comes out of their pocket, it makes no difference to Cummins. Cummins doesn't pump out all their parts within the company. They outsource.



I mean I realize for obvious reasons, a replaceable surface makes much more sense if you needed to replace a part rather than scrap an entire piece, rebuild or not. The main idea was to have a uniform surface that was replaceable if there was a problem. But IMO, the mass production costs just also made alot more sense.







But I don't see why you couldn't eliminate the bearings, then again I'm no engineer. It makes sense to me to machine the tolerances closer and still have the oil grooves, it would do the same thing. But it also makes sense to me for the purposes of production and, if neccessary, replacement, to utilize the bearing.

rotarygod

Obviously there is a reason why it isn't done in street cars. I've actually thought of several potential issues why just since I posted this and why it may not be practical. For race use it could probably work good but when made in even smaller amounts it really goes back to what you said. How often do we see those fail? Why change it if it works?

heretic

Note that most aluminum OHC engines do not have cam bearings, the camshaft rides directly in the head.



For things that take significant load, such as crankshafts or eccentric shafts, you'd want to have the bearing shell there instead of just riding in the block. Bearing shells are replaceable, so you don't have to throw away an expensive casting when it gets worn out. The bearings are also sacrificial lambs, they are designed to permit debris to a small extent to embed into the bearing material instead of scarring and trashing the journal.



Many race engines use expensive chromoly cranks, which are damned hard. I've yet to see one with damaged journals. In a related story, these are often found with "hard" bearing shells, which don't embed but they also don't melt apart at high oil temps. In this case the crank is still harder than the bearing, and the crank wins.

jspecracer7

To get proper tolerances between parts (used parts) it would be easier to swap bearings than to buy new parts which are of different clearances.

Lynn E. Hanover

There is a whole wing in the engineering building devoted to just this subject. This idea works in a very narrow area of conditions. In precision measuring we have the air bearing, where a surface is supported on an air cushion, and the supported structure is self centering using the escaping air as a form of bearing in the vertical. The Bendix Endiron air table (spelling?) is such an application. Used to check roundness to small parts of a .00010". Parts of one ten thousandth of an inch



Anywhere there is a big difference in surface speeds, and, or, where lubrication is not possible, you will find an air bearing. Ever been to a dentist? He has one.



The spearical ends of pistons in slipper and piston pumps and motors is such an application. Constant speed drives in jet engines are a good example of such close measuring. Then in use at thousands of pounds per square inch, the spherical ***** run in cups (slippers) of about the same material.



In the old ford V-8s the rod bearing on each throw is a single for both rods. The rods run on the back of the bearing which is free to spin on the crank. The side facing the crank has the lead-indium bearing surface.

The back of the bearing runs on a tin or zink plated steel shell. Probably to prevent corrosion. The steel rods are polished and run on the steel bearing shell. Same system for old Harley-Davidsons and big V-12 aircraft engines from WWII. The 1650 inch RR Merlins and 1710 inch Allisons. So it works right up to some high HP applications.



This bearingless idea occurs to every new engineer, once he/she sees how much bearings cost, and how many returns they get for failed bearings. So the bearingless idea keeps turning up in misapplications year after year. The first thought that "if there were no bearing it could not fail" has a magnetic attraction.



The answer is design it so no bearing is required, and do your research.



Go back in time to see how this problem was cured a thousand times before you came along. Look at clocks from the 1700s. Beautifull machines. Made with the most ghastly poor equipment you can imagine. Look at the Babich Difference engine.

A clock like machine that could do Algorithms. A real mechanical computer.



So bearingless machines work well, and some do it in very high stress applications. And are used exclusively in low stress applications.



Do not include removeable vice non-removeable bearings in the argument. Cams in aluminum heads are bearinged. The bearing is the whole head.



Years ago we decided to race a GT-6. What a great idea. A Spifire with a roof and a real tractor motor. Four main bearings in a 6 cylinder engine. And guess what I didn't know about the British? They are stuck in the 1920s. The GT-6 has no cam bearings. A case hardened cast iron cam runs in a cast iron block.

Bang...there goes the first engine when a piece of the broken camshaft falls into the blender.

When the guy came to dig the footers for my new shop, his British-Leyland tractor had a GT-6 engine.



One of the plus factors for removeable bearings is that oil clearances are quite large and insensitive to variation. So rather clumsey, old, poorly maintained production machinery can be used year in year out to produce a satisfactory product. The bearing clearance of .0035" is very generous. In hydraulics you might see .00014" to .00016" as a good tight tolorance. Here you are select fitting, rather than grinding to the tolorance.



So, when you look at the bearing that is made from flat sheet stock with a puzzle lock and rolled into a tube, there is more worry about the crush fit of the OD than there is the finish (installed) ID when pushed into a tube (stationary gear or rotor) of questionable tolorance.



Good clear thinking. An idea that is used every day around the world.





Lynn E. Hanover

guitarjunkie28

everytime i feel smart, all i gotta do is listen to fred and lynn...

puts me in place every time!