Even a 45mm/L66 shouldn't be able to penetrate that front plate from 650m with APBC or APCR.
Actually, this may be a jump ball. Running this through Okun's homogenous armor program the 45mm doesn't penetrate. But running it through the face hard armor penetration program has it a partial pentration with part of the body penetrating but breaking up.
I suppose my question is "How often does the 'face hard' program predict ricochets to occur?". It's simply impossible to predict 'a partial pentration with part of the body penetrating but breaking up' as an outcome. I assume you ran APHE and not APCR.
Basically, if I were modeling this, the psuedo-code would be...
Where is the hit located
What is the obliquity
Run ricochet routine
If No ricochet
then run penetration routine
If penetration, then run damage routine...etc.
An interesting aspect of the whole event is the kinematics regarding the translational movements, rotational movements and the actual time lapse of the event. The projectile is spinning at a tremendous rate. Something on the order of 10,000 RPM. Given the weight of a 7.62cm APHE (12 pounds?), this is a lot of energy involved. I've read that rifled weapons puts about 1/6th of the energy of the gunpowder into rotation of the projectile, the rest going into translational velocity and friction. In most cases, such as a projectile striking armor at an angle of 60 degrees, the amount of rotation the projectile experiences during the penetration event is insignificant. But when striking a highly sloped plate, the event is much longer and the rotation does come into play. As an example of time, we are talking on the order of a tenth of a millisecond (100 microseconds).
My take on the 'flat nosed' projectile striking a highly angled plate is that the contact point is not only small, almost a 'stripe' of contact is being made, but the contact is being worn while the actual armor being 'worked' is being 'renewed' as the projectile slides along the plate. When striking a more 'vertical' plate, even something like 45 degrees, the rest of the projectile's frontal area is rapidly brought into play. Something like 10 microseconds or less. The rotation of the projectile during such a small amount of time might be a degree or so.
But when a projectile is 'sliding' along a plate, attempting to dig in, the projectile rotation does come into play. Especially if the initial projectile contact area has become worn or broken. The reason being that the new projectile contact area (from rotation) will ride up as it replaces the worn area! This will be constantly happening as the projectile rides along the plate. Also, the rotational forces are effected by loss of material in the symmetric projectile. If a substantial piece comes off, as these Soviet designs appeared to be designed for, then the projectile becomes unstable as far as translational movement and being pointed in the direction of travel. Also, the projectile will start to rotate (yaw). An interesting point is that one study showed that faster rotating bodies actually promoted ricochet.
The projectile is reduced to striking the plate and either cracking it through shock or another failure mechanism.