So the jackhammer has a cycle where the hammer accelerates out to its maximum position, then returns. That is one cycle. You are saying there are 50 cycles each second.
There are 3 scenarios here.
1. You sample at less than 50 Hz. The accelerometer will be sampled at a different point on the jackhammer's cycle each time. You will not be able to make any conclusion here, except that the G force is that much as some point on the jackhammers cycle.
2. You sample at 50Hz. If your sample rate and the jackahammer's are both exactly matched, then you will sample the acceleration at the same point each jackhammer cycle. You will be able to conclude what the G force is at only that point. But you could also determine the mean and standard deviation of that G force since you will have many samples at the same point.
3. You sample at something quite a bit greater than 50Hz, like 100 times greater. Then you will sample the accelerometer at 100 different points along the jackahmmer's cycle. You will be able to tell the point of greatest G force, and the point of least G force. You will be able to plot the G force against the displacement of the hammer in its cycle.
You probably need to do something like 3 to get any useful data. I don't know what kind of data logger you are looking at, but something running at kHz is still running relatively very slowly.