[TaKid455]

Prewording - by no means am I proclaiming to be an expert in this field or advising fully what's required or involved. Instead, I am relaying information I have been informed, researched, & studied (to some degree). The topic is open for discussion for those who also have input in this area.

"Are you referring to the metallurgy of the metal that his SBC is made of vs a late model engine"
In part, yes. There are other areas also.

There is the metallurgy component(s) (Casting porosity, material composition (i.e. amount of X compound) , thickness, density, ect)

From there, there is the bore spacing. Siamese or individual cylinders.

Frequencies, Vibrations and Harmonics. Every object has them. Based on the above, you need to find the values/ properties of those naturally occuring. A good bit of Differential Equations. While it has been a little since I studied this area,

Next step is to find locations and find what values are of all components at those points. This sets a baseline for the sensor output.

Once that is known, you get into the dynamic component. This is where one looks to determine at what vibration level 'knock' truly occurs. Also, the various frequencies & vibrations different combinations. This is why it is necessary to adjust knock sensitivity on some modified engines.


I looked into this topic a few years back for personal Pontiac 24x conversion project. Including sitting down with an GM / Dephi engineer and going over the idea. Based on info received and some research, I paused and decided not to dig deeper to the finer details as for the intended project mainly based on what all it would take to find out what 'knock' valves would be. Yes, one could figure it out. But I would want to have imperial data throughout and not make assumptions.

Using the guess & check method
We could put a given sensor in a block. At what point would we base 'knock' values. One could evaluate a handful of factory designed applications and record data. Then take sensors for respected blocks and insert into other known blocks and record data. Then compare. Perhaps at this point, one could take tested sensors and place into experimental block and see how that data compares.

It would be an interesting experiment I'd be up dig into