4 Cylinder Inline Engine

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I have been trying to make the original crank work with a center main bearing but every thing I try just makes it worse. Those pieces will go into the "Too good to throw away" bin and I will start over on a new crank shaft. The new crank will not have needle rod bearings as originally designed but bronze split bushings and it will be flat not round.
Some of the new pieces for the crank.
IMG_3135.JPG

I wanted to make sure that I had enough room for the center bearing so I made and installed the center support.
It will be a close but it does fit and the throws line up with the pistons.
IMG_3138.JPG

Thanks for looking
Ray
 
Thanks K2

I assembled the crank pieces last night with Loc-Tite and left them to set over night.
I then installed tapered pins at all stress points and put it in the engine block. As expected it was a perfect fit but I still had not cut out the main shaft to allow for the rods.
IMG_3142.JPG

Cutting out the sections for the connecting rods.
IMG_3146.JPG

Now it should stay straight after cutting but you never know.

I didn't get a picture of it installed in the block but it did stay straight and runs smooth and true.
Now for some paint.

Thanks for looking
Ray
 
Excellent! - Hope it resolves the #57 issue! If my Math is right the crank should be 8 times stiffer? - so the out-of kilter should occur at about 2.8 times the previous 3000rpm? = 8400rpm?
Not that I think you'll get there with automatic intake valves.
Be interesting to hear your results in due course.
An excellent thread!
K2.
 
Excellent! - Hope it resolves the #57 issue! If my Math is right the crank should be 8 times stiffer? - so the out-of kilter should occur at about 2.8 times the previous 3000rpm? = 8400rpm?
Not that I think you'll get there with automatic intake valves.
Be interesting to hear your results in due course.
An excellent thread!
K2.
I think is even a bit more than your estimation.
First, 2 bearings, due to internal play, do not take any bending load (torque); central bearing is much evenly loaded on both sides and takes much more pure radial compression. And, as bonus, the other side of crankshaft (if we judge one half), supports bending load on that specific half we are analyzing. I haven't gone that far as for a mathematical model and I don't think I could remind all formula to do it :).
 
Well I don't understand all of what has been said but I am sure the crank is much stronger now.
Thanks for the comments.

I have been practicing my tubing bending skills today.
The local hardware store has some thin wall 5/16" brass coated tubing 3 feet long for $7.00. Now I have tried to bend this tubing before and it was a total disaster until I started using this stuff. Cerrobend is a low temp metal (melts at 158 degrees Fahrenheit) is poured into the tubing before bending. I put it in a pan of water on the stove and when it is melted I spoon it in the tube. It is heaver than water so the water is displaced and when the tube is full take it out and let it cool. When you are done with your bends just heat the tube and the Cerrobend just runs out with no cleanup and It can be reused over and over again.

Exhaust pipes
IMG_3156.JPG

Thanks for looking
Ray
 
Well done Ray - I must buy some Cerrobend now! - Thanks for the tutorial!
On stiffness of beams - I didn't do the "complex" 3 point supported beam, just the simple 2 point supported model
"The effective stiffness of simply supported beam is K=3EI/L^3."
Half the length (between supports) is all I considered from th 2 bearing to half the 3 bearing crank, which led my mental arithmetic to 1 over 1/2 cubed = 8 so "8 times stiffer", then the out-of balance vibration is related to the square root of 8 => 2.4 times rpm for "the same vibration as previously".... (I think? - Napier, "help" please? - I am guessing a bit here...).
BUT Napier-Deltic is correct that the 3-bearing arrangement is better than 8 times stiffer than the 2 bearing crank due to one half "stiffening" the other.. so the centre bearing is not a "pin" support, but a "partially restrained" support to crank stiffness....
But as each half is subjected to the equal and opposite loading, when rotating dynamically, then the centre may not be as well supported as Napier suggests, so the rotating "out-of-balance" bending forces are much more complex than my education...
The over-simplification I did was really just a mental arithmetic assessment of a "worse case" scenario, to demonstrate how much improvement the centre bearing should make.
I'm sure there will be a "3 bearing crank" design tutorial somewhere on the web... for those really interested. - I'm just too busy right now. Sorry.
This needs a "Proper Engineer" to explain better than my mumblings...
K2
 
Yes K2, that-s what I thought; on static, forces on both halves being on the same magnitude (not equal) looks like central bearing acts partly like a fixed attachment point for each one half. And I said (and meant) just a bit more stiffness. Of course, my "simulation" was also entirely mental and making a lot of simplifying assumptions.
My consideration, K2 and good luck Ray!

P.S. K2, thanks for the info related to resonance vs RPM vs stiffness! I have less primary knowledge about dynamic loads. Selecting te right speed when turning on lathe -to avoid vibration defects- is still a pain for me.
 
Last edited:
The calculation for a beam supported at both ens is assuming pin jointed at the ends. The 3 bearing crank acts the same for each half really.... IMHO?
K2
F is resultant of forces acting on analyzed half of crankshaft.
F1 is resultant of forces acting on the other half of crankshaft.
Situation is mirrored on either halves.
1713694606760.png

This was my judgement.
Cheers!
 
Last edited:
From that, I should calculate it as a pin jointed outer ènd and stiff supported middle bearing and calculate stiffness for half the crank length.
There is a notion that the whole length of shaft to resist bending is not the distance between bearing centres, but the combined length of the neutral axis for half the first bearing journal to neutral axis of first web, along that neutral axis to big-end journal neutral axis, along that to next web, then to next big-end, along to crank web then to mid-point of middle bearing....
Fun to calculate! Not hard.
K2
 
I guess, but not by me :) ! Last serious calculations (excluding +,-,x,: ) I have done maybe 30 years ago...
Please let me return to my cave! And I am not ironic.
Any knight needs an ideal to continue to fight ; be it Holy Grail, crusade, wealth or love...or other.
I have already defined mine and I will be much happier when I will finish -for instance- my first IC engine...if!
 
I guess, but not by me :) ! Last serious calculations (excluding +,-,x,: ) I have done maybe 30 years ago...
Please let me return to my cave! And I am not ironic.
Any knight needs an ideal to continue to fight ; be it Holy Grail, crusade, wealth or love...or other.
I have already defined mine and I will be much happier when I will finish -for instance- my first IC engine...if!
Hang on. You live in a cave, and you are a knight? You don't happen to dress as a bat and go out to catch criminals at night do you?

Jokers aside, I'm enjoying the details of this discussion. Especially as a 'mug' who tends to just throw FEM at problems like this.
 
Maybe last templar, half knight half monk; and living in a cave :) .
As I think the problem is much more complex (we were discussing here just one of the most simplified models, but if you want to go deeper, you notice it is half meaningless, unless you take in consideration other factors also), a modelling program would be of great help.
Yes, I know, we depend more and more of those damned black boxes with silicon chips inside...
 
Last edited:
Nerd, You are no " 'mug' who tends to just throw FEM at problems like this".
If you can do that you know MUCH more about mechanics than I shall ever expect to learn.
I haven' even got to "writing the application to enter", never mind "the start-line", on that one!
FEM = Flippin' Extraordinary Maths?
K2
 
You guys are talking way over my head but it is interesting.


I am working on the fuel system now and I am going to try an up draft carb configuration. All of my multi piston engines have one cylinder that doesn't work as hard as the others. Some of the cylinders are too rich and some are too lean so tuning for the best running can be difficult. This engine will have the most balanced routing I can make so we will see if this helps.

Intake box.
IMG_3160.JPG

And installed on the engine.
IMG_3162.JPG

No sharp 90 degree bends and they all should have an equal share of the fuel flow.

Thanks for looking
Ray
 
Hi Ray,
I worked in a car company - the guy who did the inlet manifold work sat next to me... a carb into a central Plenum. from which into the mid-point of a traditional "1 long pipe with 4 branches" He laboured for months, with supporting visitors from the main design group, and they "tweaked" the shapes, grooves, little air-guide baffles and shapes inside the castings to "encourage" fuel to follow the air-flow turning into the inner cylinders, instead of going straight-on to the outer cylinders! To get it right at full throttle meant it would never be right at anything less.
It is no wonder the fuel injector per cylinder made such a huge improvement to power across the whole range, smoothness, good low emissions (with positive feedback constantly tweaking fuelling), etc.
Carbs are Complex, Highly Sophisticated crude devices, that then dump a fuel-air mix into a single manifold that proceeds to separate the fuel droplets from the air so each cylinder gets a different mixture! Air and vapour go around corners, but droplets carry-on....
The Manifold "flow expert" explained all the problems to a group of us and "why" it could never be right for all engine conditions, just a crude average at any point... - but generally at its best for max power (necessarily). - Unless "fuel economy around town" was the sales USP... in which case a different version tuned for that condition was used. - Depended on the "legislation test" for the market really. - And everyone got the optimum version, so all markets just had one manifold.
It was no wonder the Racing guys could always get a better manifold for racing! - Or used 4 carbs for 4 cylinders!
GOOD LUCK tuning your design. 4 "equal" pipes from the plenum should work...
A guy at the local Model Engineering club ended up - after 3 previous manifolds - using a plenum to 2 equal curved pipes, each then dividing into 2 pipes to 1 & 2, and 3 & 4, for a 1,3,4,2 engine. But that wasn't perfect, it just "ran OK". So each cylinder "saw" the same set of curves and lengths from plenum to cylinder head. - Just "opposite hands", which didn't vary the intake flow.
Ho-hum!
K2
 
Last edited:

Latest posts

Back
Top