Reduce engine size.

[Ken I]

minh-thanh - yes I'm an engineer and I have studied things such as strength of materials and structural design plus years of experience in breaking things.

So while it may be obvious to me it is not always so to others - I've also had the salutary experience of lecturing to students which maybe helps - I can only suggest you start with the basics of strength of materials (plenty of stuff on line) starting with the basics of stress and strain, Youngs Modulus of elasticity, sectional modulus of structural element etc. etc.
If you are interested in learning you need to start at the beginning - please, I'm not being condescending here but there is no way you can start to understand more complex issues like Euler-Rankine crippling loads, principal stresses and strains etc. etc. the more esoteric it gets, the more you need the background.

Having said all that, I almost never do any calculations in the real world anymore - I just use a thumb-suck from experience and if I'm really concerned chuck in to a 3D stress simulation - Like where risk to life is involved.
Nowadays almost all that stuff is done with software - but it sure helps to have the basics under your belt.

Regards, Ken I

 

[minh-thanh]

HI K2 !
Thank you very much
I can translate it with google translate
As you know in my other thread about crankshaft fracture
so i want to know more...that helps me in some test
With a 1-cylinder engine I'm not too worried about it, but with a multi-cylinder engine it becomes important.

 

[minh-thanh]

Hi Ken I !

So without doing any calculation of actual stress then an 8mm vs a 6.9 is (8^4) / (6.9^4) = 1.8070 or 80.7% stronger.

Just being aware of the ratios is all you mostly need.

Regards, Ken I

To understand correctly
If my shaft is 5 then I increase to 8
8^4 / 5^4 = 6.5536
So the shaft will be 6.55 time stronger right ?

 

[Ken I]

Hi Ken I !


To understand correctly
If my axis is 5 then I increase to 8
8^4 / 5^4 = 6.5536
So the shaft will be 6.55 time stronger right ?

Absolutely correct !
Huge differences for small changes because of ^4
Regards, Ken I

 

[Steamchick]

Also, the other way to appreciate how size changes things, is to look at the reciprocal of you 6.55 and consider that you have reduced the stress to just over 15% of what it was previously.

BUT, as Ken 1 explained, until you really appreciate how stress is compounded by various things, you cannot properly design just by a single number (e.g. torsional stress). You must consider all the factors, Torsion, bending, stress concentration, fatigue, etc.
Everything breaks when the stress is too high, but in the real world, Design Engineers study how long it takes for things to fail, with the compounded stress from the "whole?" loading on the job (In your case a rotation beam) and how many cycles hey need it to survive without failure. - e.g. a thousand, 1 million, 10 million, 100million, 1000 million, or 1million million (The English Billion)... cycles.
We start with the stress at which the material ceases to be elastic, then de-rate the permissible stress by many factors, to get an acceptable "maximum permissible working stress" - then we design to NEVER EXCEED the maximum permissible working stress. and we often have "factors of SAFETY built-in on top of all the others...
So we end up with something like:
{Stress from bending, torsion, etc, x stress concentration factors x safety factors x fatigue factor x temperature factors x corrosion factors x every other factor (material ageing in the case of many materials!)}, MUST BE LESS THAN < the maximum permissible stress for the material.
I hope this is not too complicated to translate.
Ken 1 would have made it seem simpler.
Regards,
K2

 

[TonyM]

I appreciate the technical aspects of Kens1's and other posts but for me what is important in scaling small engines(in my case only single cylinder) is the aesthetics. I want it to look like the original. I want it to run reliably as a display model and not required to perform at scale power. Therefore the design of the internals are a compromise to match the scale outside. As long as the material selection and the fundamentals are correct then their accuracy to scale is less important. Where moving parts can be seen as in an open crank engine then scale is more important to me. Very often old engines were way ''over engineered'' anyway.

 

[minh-thanh]

I appreciate the technical aspects of Kens1's and other posts but for me what is important in scaling small engines(in my case only single cylinder) is the aesthetics. I want it to look like the original. I want it to run reliably as a display model and not required to perform at scale power. Therefore the design of the internals are a compromise to match the scale outside. As long as the material selection and the fundamentals are correct then their accuracy to scale is less important. Where moving parts can be seen as in an open crank engine then scale is more important to me. Very often old engines were way ''over engineered'' anyway.

Hi TonyM !
I completely agree with you on this.
As you can see I am building a diesel engine with 4 cylinders
If it's a 2 cylinder engine I probably wouldn't be too worried about it
With a compression ratio of about 20 : 1 , the pressure, torsional load... on the crankshaft at the position of cylinders 3 and 4 is very large.
Although with the crankshaft of the I4 I increased the size of the rod journals 1.43 times larger and made it about 1/2 shorter as well as making the piston stroke about 6mm shorter than in a 1-cylinder engine, but actually I still worry about the crankshaft and it's been my concern since I designed the engine

 

[minh-thanh]

Hi Ken I
Thank you very much .
Hi K2 !

I hope this is not too complicated to translate.
K2

Google translate + me = can understand

 

[Steamchick]

HiKen 1: Your comments:
"I almost never do any calculations in the real world anymore - I just use a thumb-suck from experience and if I'm really concerned chuck in to a 3D stress simulation - Like where risk to life is involved.
Nowadays almost all that stuff is done with software - but it sure helps to have the basics under your belt."

I understand where you are coming from - a lifetime of experience, education, practice and theory. But a part of all modelling (for many) is "making it work and function as it should". The necessary calculations, if omitted, can lead to "small disasters", as Minh Thanh has had with his crankshaft. While not earth shattering, the individual modeller with the problem has suffered some disappointment, then gained some knowledge (hopefully from our experience and expertise?) and this is them gaining their experience.
Personally, I don't stress the components of many parts either, just pressure steam boilers, etc., but use that "thumb" that is a useful rule... Then when it goes wrong I do the sums.
Actually, I enjoy doing the sums as it gives me that warm feeling of confidence that my "thumb" was right, or relief in that I had the sense to do the sums and not rely on my "thumb". Just like transferring dimensions from drawings to settings and cuts on the machines, the task is very satisfying when done properly, and it works out right first time.
So I encourage everyone to "do the sums" as a part of their design work, because it saves the "disappointments" later.
As usual, I try to "advise the right way" in print, not necessarily "follow my stupid errors".
Thanks for your advice on this topic, I enjoy your input.
K2

 

[Ken I]

Steamchick - "feeling of confidence that my thumb was right" even that can set you up for a hard landing - I couple of years back I designed a machine with a conveyor of two hollow pin roller chains with 20mm diameter rods bolted between them with M4 cap screws (all that would fit) - since it went around an idler sprocket rather rapidly (angular acceleration of 47 Radians per second per second, I was concerned about the torque reaction - so I did a quick calculation and figured it to be about what my thumb told me (±4Nm) and forgot about it.
On its first run all the capscrews sheared off on the one side and came undone on the other - did that twice with the same result.
Following Einstein's caveat on repeating the same experiment, expecting a different outcome is the definition of madness - I figured I must have under-estimated the torque reaction.
Back to the calculation - polar moment times angular acceleration squared - I'd forgotten to square it so the physical stresses were 47 times worse than my estimate. Aaaarrrgghhhh!
The rules of physics are non-negotiable.
Fortunately by using two M4 capscrews at the 1/2" chain pitch into the (non-rotating) rollers solved the problem (three systems running for two years - so its fixed).
Beware of finding what you were looking for - it's called "confirmation bias".

Regards, Ken I

 

[Steamchick]

Thanks Ken 1. A salutary lesson, and that is what makes "experience".
Been there, done it, forgotten the lesson...
K2