Cardan gear vs. slider crank in IC engines

[unclematt]

https://oa.doria.fi/bitstream/handle/10024/36120/isbn9789522145345.pdf

This paper compares the slider crank to the Cardan gear when used in internal combustion engines and was written just last year. The conclusion states that a cardan gear setup in IC engines would be superior to that of cranks sliders, which are almost universally used now. Considering the benefits offered by a hypocycloidal system, such as no side forces on the piston (less friction), closer possible tolerances between cylinder and pistion, less energy lost to blowdown, and longer stroke, this thread is devoted to any thoughts or ideas you may have on the subject.

Please check out the paper linked above to see all about this design if you are not familiar with it already.

I have especially been thinking about a "rollerized" cardan gear design using silicon nitride roller bearings to handle the loads and operating environment. This would also reduce the need for lubrication dramatically. I have also been thinking about using this design in a diesel engine. I am trying to evaluate and quantify the of forces the silicon nitride bearings would have to endure. Can anyone give me an idea of the forces that a connecting rod bearing experiences in a normal diesel engine? Both peak and continuous?

This will help me formulate an approach to deciding on the crank journal element diameter if a roller bearing was used instead of an oiled sleeve bearing. In other words, at the connecting rod, I'm trying to figure out the minimum practical inner diameter of the roller bearing given the use of silicon nitride bearings. Any help would be appreciated.

 

[MikeHalloran]

The paper didn't cover manufacturing cost.

A quick scan leads me to believe it doesn't cover weight and dynamics of real parts, either... which is where you're starting.

Slider- crank is also 'shorter'. Deck height is a major packaging consideration in automobiles.

You might be able to sell the idea in engines that already have crossheads anyway... but the added complexity doesn't help.

Okay, next question; How do you make a multicylinder engine with it?



Mike Halloran
Pembroke Pines, FL, USA

 

[unclematt]

Well, you could use a boxer configuration, or an OPOC design, and there are a couple of designs at

http://www.wiseman.com.

The Wiseman site claims you can use counterweights to create hypocycloidal engines with as many cylinders as you like. I'm not saying they are the final word on the subject, far from it. But there are various configurations that allow you to have perfectly balanced rotational components and still utilize hypocycloidal drive mechanisms. Yes, there are a few more moving parts, but the positive tradeoffs are worth it in my humble assesment.

I appreciated this paper for the level of detail they went to about ideal design constraints, and used some of their stuff to work on my own design. But I need to figure out the combustion forces that turbo direct injection diesel motors experience at the connecting rod bearings on the crankshaft. Anyone know or have the relevant data?

 

[ivymike]

You want the forces at the big-end bearings? You can come pretty close with an excel spreadsheet calculation, but to do it well you need fancier tool. Get yourself a cylinder pressure trace, and then calculate force at the small end of the rod due to pressure and inertia of everything above. Calculate the inertia forces acting on the rod itself to make it move as it must. Then add the two (x&y separately) and figure out your reaction at the lower end.

If you just want a rough order of magnitude estimate, take 2500psi and 130mm bore diameter (for a truck engine) and do the math. hint: the inertia forces subtract from the pressure forces on the higher-pressure portion of the firing stroke.

 

[GregLocock]

What problem are you actually trying to solve by going for these complex mechanisms?

Can you work out the conrod forces if you have an IMEP vs deg graph?




Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[unclematt]

Thanks for the input on figuring out those forces. After doing the numbers using your inputs, I am getting approximately 49,000 psi. Does that sound like the upper limit of the forces that occur at the big end?

As far as the problem I am trying to solve, I want to combine the surface area and configuration of the combustion chamber of piston and cylinder with the smoothness, balance, and efficiency of a hypocycloidal linear drive mechanism. Without it turning out to look like a steam engine from a century ago.

Does anyone know of a rollerized diesel engine online anywhere? Where they used roller bearings instead of oiled sleeves? Please post if you do.

 

[MikeHalloran]

Now that you've got representative loads, you should start looking at roller bearing catalogs. ... you will be disappointed.



Mike Halloran
Pembroke Pines, FL, USA

 

[ivymike]

hmmm... I wouldn't have used PSI to express a force. Is the figure you're quoting force/(projected bearing area)?

I've never seen a roller bearing turbodiesel. I can't imagine that an assembled crank would be able to handle the loads.

 

[unclematt]

And thats what I was trying to determine - if a cardan gear "built up" crankshaft could handle the loads of diesel engine operation. I am trying to figure out if commercially available bearings can withstand the pulses and loads, and if the general design I am working on is robust enough as well.

The built up crank I had in mind is not a pressed together one-piece crankshaft, but uses rollerized elements to allow each hypocycloid crank element to rotate in the opposite direction of any adjoining crank elements, if any. The rollerized elements take both engine loads and force the cardan gear parts to follow a set path, and prevents the teath of the gears from disengaging or getting out of proper timing/alignment. I am still trying to decide if its worth pursuing.

 

[unclematt]

By the way, Timken has split shell bearings that they have used to rollerize crank slide motors with pretty good success.

http://engineeringtv.com/blogs/etv/archive/2007/05/31/episode-42-rollerized-engines.aspx

I don't even think the bearing elements are silicon nitride, as I would like to use. So I am hoping the bearing upgrade adds some strength to my design, an dreduces lubrication requirements.

 

[MikeHalloran]

I missed the video where it actually ran as an engine in a truck for a million miles...



Mike Halloran
Pembroke Pines, FL, USA

 

[unclematt]

Most of the motors on the road today don't run in trucks for a million miles...

Look, I am just having fun with this and exploring some ideas. I'm not saying what I am doing will introduce anything new in the world of engines, and I am sceptical as anyone else. If shown my concept is fatally flawed I will walk away and try to find something that isn't. I still think this idea is worth exploring, and will learn while doing so.

Hypocycloidal cranks offer increased torque and efficiency according to the references I have posted, and I am just trying to take advantage of that. Please feel free to suggest alternative methods of hypocycloidal design, as I am always interested. As a matter of fact, I have a version that uses oiled sleeve bearings like all diesels use today, but I came across sources that showed rollerized engines were more efficient, and wanted to add that efficiency gain to my design. For example: small rotary engines like the Aixro use roller bearings at high RPM to create 50 hp in a little over 30 pounds. But if its a dead end for what I am suggesting, then so be it, but I still haven't seen enough data to come to a conclusion.

 

[MikeHalloran]

Truck engines are expected to run for a million miles, but I was using a little hyperbole. I found no evidence that the Timken 'engine' has run for even one mile. Granted, outboard engines have used needle roller bearings for a long time, but they're not Diesels.

I was trying also to make another point. It has been my experience that projects containing more than one leading/ edge/ breakthrough/ uncertain technology are doomed from the outset, because you've got two whole sets of development problems to solve, and investors run out of patience shortly after the first set is defeated. .. or before that.

I.e., fine, go ahead and make a hypocycloidal engine, but don't dilute your efforts by also trying to do something else that's 'different'. You'll have trouble enough.





Mike Halloran
Pembroke Pines, FL, USA

 

[unclematt]

Indeed, I have yet to find any evidence of any rollerized diesel engines at all. However, I have done quite a bit of research on silicon nitride bearings lately, and thought they might be able to make a rollerized diesel a possibility. The rolling elements are much harder and lighter, they run cooler (though they can resist very high temperatures),and they don't require much more lubrication than grease. So maybe someone with the resources should give this a try.

Your comments about incorporating more than a single untested concept into my overall design are well taken. I am still interested in following the cardan gear approach a bit farther though.

 

[patprimmer]

Two stroke petrol engines use roller bearings as the tend to be high rpm low torque engines and as fuel runs through the crankcase, they have very marginal lubrication.
Large cylinder, especially diesel engines use slipper bearings with a pressurised oil supply for good reasons.

A slipper bearing has a much larger bearing area to accommodate the shock forces associate with the combustion process in a diesel engine.

In one phase of development of racing air cooled VW beetle engine roller bearing cranks became popular as the VW std oil system was not up to the task in a high power high rpm engine that transferred considerable heat to he oil.

As power was further increased by increasing cylinder pressures and therefore torque and as oils improved and as the need to use a large oil cooler became recognised, the swing was back to slipper bearings to accomodae the higher loads.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[GregLocock]

/OT/ is slipper bearing a common name for a plain journal? The reason I ask is double slipper/triple slipper bearings.




Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[patprimmer]

Greg

I have always understood slipper bearings to be the type of plain bearing that uses replaceable half shells. I don't know why I think this or were I learned it.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[GregLocock]

Sounds right, it's just not a phrase I've come across before.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MikeHalloran]

I've heard it before, and understand it as Pat does, but I can't remember the source right now.


Mike Halloran
Pembroke Pines, FL, USA

 

[unclematt]

An issue I am encountering while tryng to work on my cardan gear design is that you need a minimum crank element diameter to transmit all the torque and horsepower without failing or distorting. However, since the ring gear is twice the diameter of the pinion gear, this forces a relatively long stroke since the pinion gear cannot be less in diameter than the crank shaft (or a weak area is created). So if you have a minimum crankshaft diameter of lets say 50mm (hypothetical), that causes the ring gear to be 100mm in diameter, which makes the stroke 100mm AT LEAST.

One of the benefits of the cardan gear crank system is that it allows longer strokes, because the rod is moving in a linear fashion and cannot strike the block. But of course there is such a thing as TOO long a stroke, especially in relationship to the cylinder diameter. While I am certainly trying to increase efficiency by reducing blowdown, I don't want to end up with a stroke that is too long. So I am working backwards from the compression ratios I am interested in achieving (18.5 to 20 to 1) to the sizing of the cylinder and stroke. Comments and input, as always, are welcome.