Needle bearings in a model diesel ?

[Lloyd-ss]

I've got a thread going on a 56cc 2 stroke diesel that I am building. The design process is odd in that I have what I consider the most difficult parts, the roots blower and the unit injector mostly finished and will now be designing the rest of the engine around those successes.
Question is the possible use of needle or roller bearings for the crank and pin to avoid some of the oil mess. I am fully aware of the many advantages of sleeve insert bearings, but needle and roller bearings are used in many engine applications, so why would they be a total bust on a low power and low speed diesel? I am researching now and am hearing mostly "no", but if sized and chosen carefully, seems like they might work. If the design is based on needle, and not a retrofit.
Thoughts? Other than "it will never work," LOL.

 

[Vietti]

I don't know for sure. Chain saws have needle bearings on both ends of the rod and live a hard life seldom failing.

I've used them in a couple low speed gas engines and so far so good. Currently building a Hoglet with a needle bearing on the rod. As you know the issue is once the bearing is on the crank, depending on the construction, you cannot replace them. Also go with largest bearing the design will allow.

I look forward to your design and outcome!

 

[GreenTwin]

I am no bearing expert, but it seems like to boils down to the pressure on the bearing face.

A ball bearing has a point contact.

A roller bearing has a line contact.

A babbitt bearing has a rather large support surface.

My dad always said that a well oiled babbitt bearing will outlast any other roller type bearing.

I think the babbitt bearing has the smallest amount of pressure for a given area.

For a model engine, I think a roller or even a ball bearing would be ok, assuming you replace it if it gets worn and starts knocking.

.

 

[Lloyd-ss]

I think I have seen needle bearings that have split cages, but they might require a decent supply of lube oil. If I want to keep the oil mess to a minimum, meaning sealed bearings, the split cage might be a problem. But there ought to be a solution.

 

[Joe]

Would the difference in hardness between the rollers and the crank be a consideration ?

John

 

[Klrskies]

Perhaps you could calculate or scale the pressures that will be exerted on surfaces that will contact the rollers and if the hardness is adequate, maybe around 60 rockwell C, to avoid deformation?

 

[Steamchick]

More "I think it will work, but really just guessing" comments from me....
The bearing designers have (or did have 30 years ago) very good design guides for their bearing applications.
I have used then with success. They take the "expertise" out of the design as all you need to do is follow their rules and sums to decide a suitable size of bearing per application (big end, small end or main).
But here are some rules I have picked-up over the ages...
Plain bearings need adequate pressure and flow (to keep the bearing cool) so have high pressure oil pumps to force adequate oil through the bearings. Needle roller bearings only need a little pressure but a lot of flow for bearing cooling. Centrifugal oil pumps are good.
On the flywheel end, the flywheel forms a gyroscopic high inertia, so is almost a fixed plane against which the crank flexes and strains the drive bearing (main bearing adjacent to the flywheel). So a single row ball race is the best here, not a needle roller. Other bearings are not so susceptible to the crank flexure. But a narrow bearing is much better than a wide one for crank flexure interaction with the needle roller. A "stiffer" journal crank is favoured = Larger diameter and narrower journal.
So decide your journal diameter, and use a needle roller just long enough for the design, to the maker's guide. Longer is not better.
Splash lubrication is probably inadequate for bearing cooling. So a drilled crank is necessary with pumped oil.
A hard crank pin is a must. "Drill" steel or something?
A Precision diameter for the needle roller to run on is necessary.
Can't remember other odd stuff...
K2

 

[Roger B]

A plain bronze bearing will probably be fine unless you are intending to really use the engine. Schillings made a number of designs with roller bearings, but these used a built up crankshaft. Split roller bearings such as these are available:

https://www.gizmomotors.com/online-store/twin-needle-bearing

 

[Steamchick]

Hi Roger, There are various ways of "killing the pig" - What does it matter when you have Pork on the plate?
But if you need to decide on "peripherals" such as oil pump, cooling system etc., then it is a good idea to explore the options and decide on such things as Crankshaft design, block design considering all the options of bearings.
Some motorcycle engines have solid cranks with plain bearings, as do most cars, I think most British singles, and Japanese engines have rollers and balls on built-up cranks as they suit the application, and peripherals such as oil pumps are designed accordingly.
So if The Gentleman wants to understand Needle Rollers, let's help him with our knowledge, and not just say something else will work, please? This is a strategic decision, so should be taken with care and the best advice available - which is why he asked.
Not to put a damper on your comment, but really to ask "What are the advantages and pitfalls to avoid of various designs?" - Can you elaborate on Bronze bearings per se? - I love to learn from others.
Cheers,
K2

 

[ajoeiam]

More "I think it will work, but really just guessing" comments from me....
The bearing designers have (or did have 30 years ago) very good design guides for their bearing applications.
I have used then with success. They take the "expertise" out of the design as all you need to do is follow their rules and sums to decide a suitable size of bearing per application (big end, small end or main).
But here are some rules I have picked-up over the ages...
Plain bearings need adequate pressure and flow (to keep the bearing cool) so have high pressure oil pumps to force adequate oil through the bearings. Needle roller bearings only need a little pressure but a lot of flow for bearing cooling. Centrifugal oil pumps are good.
On the flywheel end, the flywheel forms a gyroscopic high inertia, so is almost a fixed plane against which the crank flexes and strains the drive bearing (main bearing adjacent to the flywheel). So a single row ball race is the best here, not a needle roller. Other bearings are not so susceptible to the crank flexure. But a narrow bearing is much better than a wide one for crank flexure interaction with the needle roller. A "stiffer" journal crank is favoured = Larger diameter and narrower journal.
So decide your journal diameter, and use a needle roller just long enough for the design, to the maker's guide. Longer is not better.
Splash lubrication is probably inadequate for bearing cooling. So a drilled crank is necessary with pumped oil.
A hard crank pin is a must. "Drill" steel or something?
A Precision diameter for the needle roller to run on is necessary.
Can't remember other odd stuff...
K2

Hmmmmmm - - - - the was a rather notable outlier in the lubrication department.
Detroit diesel engines (notably the ones manufactured from some time in the late 1930s until some time in the late 1980s) used a lubrication system that was based more on flow than on pressure. Most engines rely on pressure to maintain bearing clearances (the hydrodynamic part of the bearing system(s)). Those unused to operating the Detroits would be quite surprised to see an idling engine with its oil pressure being down in the 7 to 10 psi range. Most other diesels, at least from my ken, would need a rebuild when the warm idle pressure was dropping below 35 psi (if not even higher - - - - all depending upon the manufacturer - - - -check your technical manuals).

(I understand that m'sieur Steamchick indicated 'adequate pressure and flow' - - - most rely more on pressure for indication of health rather than flow.)

HTH

 

[Roger B]

There are plenty of design considerations here. The first is that this is an experimental engine that may or may not work. The critical parts have already been identified as the fuel injection system and the blower. Small size diesel combustion is also a consideration.

The engine as I understand will only be run for demonstration and will not be seriously used. In this case I would make the ‘standard’ engine parts as simple as possible and save the complications for the ‘difficult’ parts.

Needle and roller bearings have benefits for crankcase compression petrol two strokes where they can be simply lubricated by the oil mist. Construction of the crankshaft and big end is more difficult, requiring a built up crankshaft or split roller bearings, both of which are possible as I said.

As this engine does not have ‘petroil’ lubrication something different is required. The simplest would be a squirt from an oil can now and then. Some form of splash lubrication would also work. Pumped lubrication would be better, this could just be a spray nozzle pointed at the big end. The ultimate would be to drill oilways in the crankshaft and use a pumped pressure feed as in full size engines. I chose this option, although in hindsight for the amount of running so far a few squirts from an oil can would probably have worked.

If ball or roller main bearings are used it is more difficult to arrange a pumped pressure feed to the big end.

A one piece steel crankshaft with a split big end and split bronze bush has worked well in many small engines. For the sort of duty anticipated mild steel is probably enough although I have used moderate carbon steel (C45k). A standard bearing bronze such as RG7 works well.

 

[Steamchick]

Thanks Roger. Fair comment. But back to Ajoeiam's response to my "lubrication" ideas. Plain bearings have a constant small gap resisting a fed lubrication, therefore need a higher pressure to develop adequate flow to cool bearings. This is NOT the hydrodynamic pressure. The hydrodynamic pressure is developed as the bearing is loaded - typically from the combustion pressure acting on the piston, but also from dynamic acceleration loads from reciprocating masses. This hydrodynamic pressure is the bearing load divided by the acting area of the bearing and is typically much higher than the fed oil pressure. The hydrodynamic pressure tries to force the oil from the space between bearing and shaft. The fed oil pressure feeds oil when the bearing is less loaded during the rotational cycle.
Conversely, a rolling element bearing has stationary contact points, and gaps between them. The gaps are relatively huge, so pass oil very easily, so oil supplies are relatively low pressure and still have adequate flow to provide the necessary cooling of the rolling elements. I have owned Honda (4 stroke) motorcycles that run on thin oil with just a few psi from the pump through the crank and the ball and roller races. I have also had other engines with plain bearings with high (many 10s of psi).
Considering the diesel develops very high pre-combustion pressures just to ignite the fuel, then has the very high combustion pressure on top, bearing design for both crank and small end bearings is critical to the success of the engine. Both plain and rolling element bearings can achieve a good result, if designed correctly with crank pin sizes, loads, etc. But what is selected for the engine then determines the necessary lubrication scheme, not the other way around.
Hope this helps Lloyd?
K2

 

[Lloyd-ss]

Gentleman,
This is ALL very helpful and a brisk discussion covering many different methods, from many different points of view, is ever-so-nice. I had a doctor as a neighbor years ago, a super nice guy who was also handy at home. He would come over sometimes and say, "I need a Lloyd consult about a plumbing problem," or something like that. He had one person to consult with, I now have many. That is quite a luxury, and I do realize that and appreciate it.

What I am picking up is that I have many different options that will work, so long as I design the engine components around the bearing choice(s) and play by the rules of each particular bearing. I also realize that what I choose might not work, but it won't be the first time that I have had to redesign-remake a part or assembly because it didn't work. When possible I like to make a part with the possibility of rework in mind so that I don't paint myself into a corner.

Pondering, pondering.....
Lloyd

 

[Steamchick]

Hi Lloyd. Whatever you choose, we can help, but we have varying degrees of expertise.....
I am not an expert on crankshaft and bearing design, but sat next to a few..... and they passed on some gems. But I have never been the "drip under pressure"... = ex- spurt!
K2

 

[Lloyd-ss]

Hi Lloyd. Whatever you choose, we can help, but we have varying degrees of expertise.....
I am not an expert on crankshaft and bearing design, but sat next to a few..... and they passed on some gems. But I have never been the "drip under pressure"... = ex- spurt!
K2

emphasis in the quote added by Lloyd.

Steam,
How true, how true!. It is amazing what you can learn with a very few questions and a lot of listening. And that is why I am here (hear).

Lloyd

P.S. and to hopefully share a little bit of info, too.

 

[Steamchick]

Lloyd, your good question has elicited good and interesting answers. Keep asking!
K2

 

[peterl95124]

FYI on needle bearings

I tried to use needle bearings but wasn't able to for various reasons...
* they only have an external housing, no inner race, they are designed to run
directly on a shaft, the shaft has to be hardened <---
* the external shell is designed to be press fit into a *very* solid housing,
this will shrink the shell so the needles are OK for the nominal shaft diameter
they were designed for, if you plan (like me) on using loctite to hold the shell
in place then you won't be able to find any drill blank (fractional, number, letter, etc)
that matches the un-shrunk diameter of the needles, you'll need to grind one down
yourself until it fits. (One thing that I still wonder about is how thick does the housing
have to be to shrink the shell correctly, and how much does housing metal and alloy
affect this?)
after figuring this all out I gave up, didn't want to try grinding my own custom diameter
shaft, though I do have a tool post grinder for my lathe and may try it in the future,

needle bearings don't seem to be an off-the-shelf solution for model engines (where you
don't have enough housing material to withstand the press fit they are designed for).

if anyone else has contrary experience that enables use of needle bearings in model
engines I'd like to hear about it, I still think they would have their uses, especially for
blowers and superchargers (personally I'll stick with bronze bearings for crankshafts
and crankpins).

 

[Steamchick]

Hi Peter.
The only time I used a needle roller in an engine was a main bearing adjacent to the cantilever crank of an aero- engine styled 2-stroke - 1.5 cc - where I turned the crank to suit the needle roller.... it was a gentle press fit in the aluminium housing. It needed an oil seal behind it for the 2-stroke crankcase case seal. Bearing worked fine.
K2

 

[kf2qd]

Bronze bearing. Ran them in model airplane engines, high speeds, lots of hours. don't need to be very thick, don't require a lot of extra material for strength.

 

[barrti]

I have always been impressed with needle rollers as used in petroloil engines. They survive high loadings and rpm just fine. This means they only need scant lubrication especially considering the low power levels here
4stroke glow engines use journal bearings in the conrod and these rely on the blowby past the larger than usual ring gap to get some oil mist into the crankcase. The big end and small end of the conrod is drilled to allow oil into the journal bearing. The crank case is grooved to allow some the oil spray to collect there the big end actually runs in this groove and picks up enough oil apparently to oil the journal bearing. Hydrodynamic bearings btw develop their own pressure from the relative rotation of the shaft within the conrod. As long as the bearing gets enough oil supply to the bearing to cool it, it develops its own pressure wedge to prevent metal to metal contact
They are fascinating..try. Reading this link https://www.sciencedirect.com/topics/materials-science/hydrodynamic-bearing

Anyway, the 4 stroke glow engine which runs 20 percent castor oil to methanol fuel, encourages me that a two stroke Diesel may work relying on the blowby and oil mist in the crankcase. If you want more security, an occasional drip of oil fed to the groove in the crank case might help. As someone else has suggested, use the bearing manufacturers for info on shaft hardness, finish and required press fits. Loctite could help if you feel you can't get enough press fit in the rod. Don't forget, the gas two stroke engines obviously get enough press fits in their steel conrods.
When you select a rolling bearing, the main inputs to the process are speed and loading. Loading could be either guessed or estimated using some calculations based on engine balance equations. Another way is to just look at two stroke gas engines of similar size and use their dimensions.
The main bearings could be sealed or open ball bearings. 4stroke glow often use a sealed ball bearing next to the propeller and an open ball bearing next to the conrod..
These are just my initial thoughts, I hope they contribute.
Have a good look at a 4stroke glow engine if you get the chance.
If you need more info, I can probably chase it up.