mpopenker said:

       RovingPedant said: I would expect that the forward bolt stroke starts with zero initial velocity on every machine gun ever made. Did you mean zero spring load?

Not necessarily When bolt group bottoms up the spring and hits the rear of the receiver, it is usually an elastic impact, and bolt group receives some of its remaining recoil energy back, to start with some initial velocity (imagine a steel ball falling on a steel plate and bouncing back)

A steel ball falling on a steel plate reaches a velocity of zero before bouncing back.  The same is true of a machine gun bolt.  Initial velocity is zero.

mpopenker said:

For example, in a typical Kalashnikov AK assault rifle bolt group rebounds back from the rear trunnion at the initial velocity of 3-4 m/s under normal conditions, adding some significant surplus KE to the amount of energy stored in the return spring. This is one of the reasons why AKs are so reliable, but also results in an 'excessive' felt recoil. TANSTAAFL.

The AK is firing from a closed bolt, while nearly all MGs (and the Evolys in particular) are firing from an open bolt (but you already know that). That means that the first round is always fired with zero bolt initial velocity and the return spring should be designed with this point in mind.

Yes a steel ball falling onto a steel plate does reach a velocity of zero, right before gaining back some of the velocity it just imparted to the steel plate from an inelastic collision (or is this technically an elastic collision? I'm not an engineer i just play one on tv) and the restitution rate of the two materials in question.

It actually can gain back a fairly substantial amount of the energy it just dumped into whatever it hit depending on the restitution rates of the two materials in question as well as other relevant factors which I poorly understand other than knowing in an abstract way that they exist.

If we're talking about the AK and pkm as they currently exist, both designs do in fact impart extra energy to the return stroke from their collision with the rear trunnion. Whether another design would do so or not and to what degree would heavily depend upon the specifics of that design, however in these two cases Max is entirely correct.

This is one of the stealth advantages of the pkm operating mechanism with it's pull out then push forward belt feed mechanism. It actually lets you shorten the receiver quite substantially without compromising reserve power.

The chicoms make 308 non disintegrating belt pkm clones that still pull then push feed using the signature pkm claw as it is and it's not totally impossible to make a disintegrating link belt for a pull then push feed gun.

When you consider that it will probably be easier to make lightweight polymer pull out disintegrating links it's not entirely out of the realm of possibility that true next generation belt feds could use pull out links.

roguetechie said:

If we're talking about the AK and pkm as they currently exist, both designs do in fact impart extra energy to the return stroke from their collision with the rear trunnion. Whether another design would do so or not and to what degree would heavily depend upon the specifics of that design, however in these two cases Max is entirely correct.

No, Max is entirely incorrect.  Max said that the bolt starts with "some initial velocity."  

Even the PKM bolt starts with zero initial velocity.  https://youtu.be/01mrTVUJ_Bo?t=45

Shot 1 of all open bolt weapons begins with just the forward spring power of the weapons spring. 

Shots 2-200 may or may not have additional power exerted on the forward stroke

roguetechie said:

When you consider that it will probably be easier to make lightweight polymer pull out disintegrating links it's not entirely out of the realm of possibility that true next generation belt feds could use pull out links.

An important part would also be to limit the belt acceleration. The Evolys belt movement starts during the feed phase, just after pushing the round out of the link, and this seems to dramatically reduce the belt peak acceleration.

Shot #1 or shot #200, does not matter.  Initial velocity of the bolt is zero.

Shot #1 starts with the bolt held in place by the sear:  Bolt velocity = zero.

Shots #2-200 start with bolt rearward motion stopped by the receiver, at which point bolt velocity = zero.

Unless "initial velocity" is being defined as velocity of the bolt at some point other than the point where the bolt's forward movement is initiated?

Elastic.

Inelastic collisions are where the objects stick together in a lump.

stancrist said:

Unless "initial velocity" is being defined as velocity of the bolt at some point other than the point where the bolt's forward movement is initiated?

Id define what Max is talking about as "Spring +" power. 

Shot 1 = bolt traveling forward operating purely on spring power

Shot 2-200 = bolt traveling forward operating on spring power + energy of the bolt 'bouncing' off the back of the receiver 

Personally I think designing the weapon to rely purely on Spring Power, and minimizing bolt impact to improve controllability/reduce shot dispersion is the way to go. But it;s entirely believable that this "Spring +" effect of the bolt bouncing off the receiver does allow for more power for stripping rounds off the links.