Regarding the difference between aluminium and iron parts I'd start my research at the specific yield strenght (yield strenght per density). That will give you the corresponding amount of aluminium you need to sustain the same amount of force as the original iron part. Add 5% due to technological difficulties with aluminium though (larger termical elongation, etc.). That'll give you an estimate until you have better data.
Try working out the necessary lenghts and widths for some parts of the engine like conrods. Your formulas should include the terms that are necessary for the part to work and also a part that accounts for all material added for engineering reasons.
In the example of a conrod you absolutely need:
- Twice the thickness of the small end bearing shell, determined by fatigue strenght of cast iron and the amount of centrifugal force generated by the cylinder
- The diameter of the small end bearing
- Stroke
- Height of cylinder minus distance from small end bearing to cylinder surface
- Twice the thickness of the big end bearing shell
- The diameter of the big end bearing, determined by friction and thermal properties of the bearing material and physical properties of the cranbk shaft
On top of that you need some space to account for other influences you can't estimate. When you start researching real part's dimensions you should have every part broken down into such necessary and optional dimensions. Then measure both necessary and optional dimensions from real parts. If your formulas are correct you should recieve a gaussian distribution for each dimension. If not, recheck.
I hope that helps.
Vain