Two Cylinder Uniflow

[PaulR]

Another bonkers plan is emerging... a uniflow engine with four cylinders!

I'm thinking of using a single piece of brass for the cylinder block with some sort of inlet manifold on top. For simplicity, and to enable me to make all the parts in groups (and ensure they're identical):

... the eccentric 'rods' will just be shaped sheet metal with holes at each end.
... the eccentrics will be made from a single piece sliced into four, fixed with grub screws
... a forked guide might be needed for each eccentric 'rod' to stop them wandering left/right - maybe attached to the cylinder block (as shown here) or the standards or somewhere else 
... the shaft and the standards will need to be chunky as there's no central support

Here's the basic concept (ignore the random proportions!), based on the previous experiments it'll probably be extremely noisy...



Hope to make a start on this some time next week.

[Sanjay F]

I like it, another unique engine begins!   

[crueby]

I've never seen  eccentrics  used instead of crank pins on a piston. Offset can be the same, but it seems like the force transferred would be maybe half? Has anyone else  tried it?   

[PaulR]

I've never seen  eccentrics  used instead of crank pins on a piston. Offset can be the same, but it seems like the force transferred would be maybe half? Has anyone else  tried it?   

I hadn't even considered that but now I see that the force isn't being applied to 'one point', the pin, it must be weaker I guess although the energy has to go somewhere. I wouldn't know where to begin trying to calculate it though - the last time I did any mechanical engineering calculations was in 1975 (and I struggled then!).

[Charles Lamont]

Think about the friction.

[PaulR]

Just worked out the contact area between the eccentric and a 'rod' made from sheet and compared it to a big end on a crankpin. Plucked figures out of the air that are typical for the sort of sizes I use. Bizarrely the results are the same but in practice I think the eccentric would be larger and therefore create more friction.

Eccentric 12mm dia:perimeter 2*pi*r = 37.68
Contact area of 1.5mm thick sheet metal 'rod/strap': = 1.5 x 37.68 = 56.52 mm²

3mm pin perimeter: 6.28 x 1.5 = 9.42
Contact area of boss: 9.42 x 6 = 56.52 mm²

[PaulR]

The guides would also add friction but, as far as I can see, there's no need for the rod/strap to surround the eccentric - it could be an open 'fork', reducing contact area by 50%. I suppose it's the same as a cam set up?

[crueby]

I would worry less about the friction and more about the lever arm. At mid stroke, the force would be centered halfway out the cam, rather than out where the crank pin would have been. Much less torque  transfer.

[PaulR]

Hmm, I don't have any suitable brass for the cylinder block and large pieces are eye-wateringly expensive these days  I do have that pig of cast iron but I might need to do some soldering so that's probably out too. What I do have however, is almost 8 foot of 1/2" thin walled brass tube from two towel rails! I'm guessing it's drawn as I can't see any seam but even then I didn't think it would be much use. Seems I was wrong - I made a steel test piston and tried it over a five inch length and it's a fantastically smooth and air tight fit, at least as good as any bore I've been able to produce.

I figured drilling the exhaust ports would be tricky without support and when breaking through as the wall is only about 0.6mm thick so I turned down a tight-fitting dowel and used a simple widget to for a test. Absolutely no issues and the piston is still a great fit.

So, if this idea comes to fruition, I think I'll use four lengths of towel rail for the cylinders (funny old world isn't it?) and solder them carefully between plates at top and bottom. If it doesn't I'll certainly use this tube for several other engines!

[PaulR]

I would worry less about the friction and more about the lever arm. At mid stroke, the force would be centered halfway out the cam, rather than out where the crank pin would have been. Much less torque  transfer.

Thanks Chris but I can't quite get my head round that , can you clarify a bit please?

[PaulR]

Ignore that, I've just worked out what you mean.... thinking more.

[PaulR]

I think you mean as per the red arrow rather than the green?

[crueby]

I think you mean as per the red arrow rather than the green?

Exactly.

[PaulR]

Hmm, wonder if it would still run but just be a bit weedy....?

[crueby]

Hmm, wonder if it would still run but just be a bit weedy....?

One sure way to find out...  It could be an interesting experiment to try it with the eccentrics and then swap parts to a normal style crankshaft. Or, mock it up with one eccentric and one crank, and just move it by hand to feel the difference?

[astroud]

I suggest trying with one of your existing engines, try to turn it over by pushing/pulling the valve rod.

Andrew

[crueby]

I suggest trying with one of your existing engines, try to turn it over by pushing/pulling the valve rod.

Andrew

Good idea. Pick one with a large throw on the valve rod. 

[PaulR]

Thanks for the suggestions!

The engine I've just finished can be turned over by pushing/pulling the valve rod but it's certainly not as easy as using the crank or turning the flywheel which proves Chris's point perfectly. I only have one other engine with an eccentric which is very small and thus difficult to get hold of the rod to apply any force but I can move it a little (that engine is a bit tighter anyway, probably as it's been standing unused for over a decade.

I'll think about making a crank instead but that has the knock-on effect of needing more complicated con rods. For these basic, mostly experimental engines I'd like to keep the number and difficulty of parts down to avoid wasting time and materials in case the design is flawed.

[PaulR]

Oh yeah, the other thing about trying to pull/push the valve rod is I'm also having to drag the piston and rod etc which wouldn't be the case with the original concept so maybe it would be so bad. The other advantage is that the timing could be adjusted. Perhaps I'll be a bit more realistic and try it with two cylinders...

[crueby]

If you can lengthen the stroke of the cam version, that would give it more leverage while keeping the simplicity of  construction.

[PaulR]

If you can lengthen the stroke of the cam version, that would give it more leverage while keeping the simplicity of  construction.

I'm sure I could, I have 8ft of tube to play with after all 

[PaulR]

Might be better to scratch the guide fork idea which might be tricky to line up properly and make the eccentric a bit more traditional with a sheave or even just a little tab to  keep the rod on track. Just making the smaller diameter (spigot part) longer might work just as well though.

I think a two cylinder version is definitely the best place to start. Updating the thread title!

[Charles Lamont]

I would worry less about the friction and more about the lever arm. At mid stroke, the force would be centered halfway out the cam, rather than out where the crank pin would have been. Much less torque  transfer.

Oh, good grief. For the same stroke S, the lever arm is exactly the same S/2, whatever the diameter of the crankpin.

But compare the moment arm about the crankshaft at which the frictional resistance acts between pin an big end with the moment arm in the case of the eccentric and bigger end.

[crueby]

I would worry less about the friction and more about the lever arm. At mid stroke, the force would be centered halfway out the cam, rather than out where the crank pin would have been. Much less torque  transfer.

Oh, good grief. For the same stroke S, the lever arm is exactly the same S/2, whatever the diameter of the crankpin.

But compare the moment arm about the crankshaft at which the frictional resistance acts between pin an big end with the moment arm in the case of the eccentric and bigger end.

Wrong, Charlie Brown (good grief pun). The lever arm is at the middle of the crankpin. With the pin size in the normal crankshaft design range, its a very small difference. But, for the way this one is drawn, its a huge difference. Look at the picture in reply 12 again!

[internal_fire]

Wrong, Charlie Brown (good grief pun). The lever arm is at the middle of the crankpin. With the pin size in the normal crankshaft design range, its a very small difference. But, for the way this one is drawn, its a huge difference. Look at the picture in reply 12 again!

You missed the part about the same stroke. (Of course the proposal that started this tangent did not have the same stroke for the two styles.)

Gene

[crueby]

Wrong, Charlie Brown (good grief pun). The lever arm is at the middle of the crankpin. With the pin size in the normal crankshaft design range, its a very small difference. But, for the way this one is drawn, its a huge difference. Look at the picture in reply 12 again!

You missed the part about the same stroke. (Of course the proposal that started this tangent did not have the same stroke for the two styles.)

Gene

I did not miss that at all. Don't confuse stroke distance with lever length at midstroke. Build a simple mockup of both crankshaft and eccentric with equal throws, push on the conrod, and see how much torque you feel on the crankshaft. It will NOT be the same.

[Charles Lamont]

Gene, I have read the whole thread again, and I have still missed anything about two styles, let alone with different strokes.

Chris, here is a diagram. The lever length for the same stroke is the same in both cases. It is the throw of the crank or eccentric. You are right that it will not work the same, but that is because of the radius at which the friction acts. Hint: imagine with a plain bearing and then with a needle roller.

[crueby]

Gene, I have read the whole thread again, and I have still missed anything about two styles, let alone with different strokes.

Chris, here is a diagram. The lever length for the same stroke is the same in both cases. You are right that it will not work the same, but that is because of the radius at which the friction acts. Hint: imagine with a plain bearing and then with a needle roller.

Okay, I  like that diagram better and agree. The one in reply 12 had me thinking the crank center was in a different place.   And by two styles, I  read that as pin vs cam style?

[internal_fire]

Gene, I have read the whole thread again, and I have still missed anything about two styles, let alone with different strokes.

I probably did not state it very clearly. I was agreeing with you 100%.

Gene

[PaulR]

Glad you're all in agreement, shame I have only the vaguest idea of what you're all talking about 

I'll draw up a plan and put it here for comments before getting stuck in to building it.

Thanks all for the help 

[Charles Lamont]

Glad you're all in agreement, shame I have only the vaguest idea of what you're all talking about 

Well, if you want to reliably produce designs that work, I think it is an example of the kind of thing you should be trying to at least develop some feeling for. I know you are being at least partly facetious, but dismissing as 'too difficult', things that people have taken time to try to help you with may not go down too well.

But don't let that discourage you from asking questions - there are no stupid questions.   

[astroud]

If you are hesitating to make a conventional crankshaft because of concerns about the complication of the conrods consider using your normal conrods, assemble to the crank pins and then make up the crankshaft in the normal build up process. Loctite 638 will hold it all together with no need for pins. If it is a fail put the blow torch on it, disassemble and recycle the parts.

Andrew

[PaulR]

If you are hesitating to make a conventional crankshaft because of concerns about the complication of the conrods consider using your normal conrods, assemble to the crank pins and then make up the crankshaft in the normal build up process. Loctite 638 will hold it all together with no need for pins. If it is a fail put the blow torch on it, disassemble and recycle the parts.

It's not so much the complication as the extra time and materials, the increase in dimensions it would necessitate and, with the original four cylinder idea, the difficulty in getting the pins in exactly the right places (not something I've even contemplated before). But yes, having the rods 'permanently' attached is an option 

[PaulR]

...dismissing as 'too difficult', things that people have taken time to try to help you with may not go down too well.

I haven't dismissed anything as 'too difficult'. I've said I don't understand plenty of times but usually manage to figure things out.