Static and Dynamic Loading Modes

Tony Baker
September, 2003


This document has been public for just four days and I have realized that some of the theories herein are incorrect. So, I am revising it and the revision should be completed within a couple of weeks or sooner. Give this one a read and see if you can find the errors. 9/14/03

Historically, the creation of flakes has been divided into the major categories of pressure and percussion, which are derived from the motion of the knapper. These two categories are further subdivided into normal pressure, levered pressure, direct percussion and indirect percussion. Within these various categories the force application tools (hammers, pressure tools, etc.) are also different. As a result of these categories and their different tools, it has been assumed that the flakes created by these various methods should also be different. However, I suggest that the only real difference between the methods of pressure and percussion is the size of the flake.

This paper introduces a new dichotomy of flake creation using the categories of static loading and dynamic loading. Unlike pressure and percussion, static and dynamic loading modes are based on the motion of the core and not the motion of the knapper.

When force is applied to a core, but a crack does not initiate, the core deflects or bends. This deflection cannot be seen with the naked eye, but it does occur. When the force is removed, the core returns to its un-deflected shape. If the force is removed instantly, the core will vibrate just like a tuning fork and, just like a tuning fork, its vibration will dampen out and assume its un-deflected position. Additionally, each core has its own frequency and that frequency depends on its material and shape. If the force is applied to the core slowly, as is the case with pressure, and then instantly removed, the direction of the motion of the core is opposite the force and toward its un-deflected shape. This core behavior is caused by static loading. If the force is applied rapidly, and then instantly removed, as is the case with some percussion, the first motion of the core is in the direction of the force, which adds more deflection. This core behavior is the result of dynamic loading.

Two identical cores are depicted in Figure 1. They represent the instant in time when the crack initiates. Just prior to this instant, the crack had not existed and the total core had been deflecting to the right because of an increasing force to the right. At the instant in time represented in Figure 1, the crack is initiating and the force is being released from the core's side of the platform. If at this instant the velocity of the core's side of the platform is toward the left, then the core has been statically loaded. On the other hand, if the motion of the core's side of the platform continues with the flake's side of the platform or to the right then the core has been dynamically loaded. In both loading conditions, the flake creation force remains on the flake's side of the platform and continues to push the flake's side of the platform to the right.

The crack is propagated in the static loaded mode by the opposite motions of the core's side of the platform and flake's side of the platform. In a sense, these two motions are ripping the core apart as one would tear a piece of newspaper. This tearing will continue until the crack tip propagates to the surface of the core or until these two opposing motions stop. In the case where the crack tip propagates to the core surface, a feathered flake or a full-length flake is created. If the core is long enough then the tearing motion can slow down to the point that the crack stops propagating, and then the knapper's follow-through will break the flake out with a step termination.

In the dynamic loaded mode, the crack is again propagated by the separating of the core's side of the platform from the flake's side of the platform. But, since they are both moving in the same direction, this separation is caused by the core's side of the platform slowing down in relation to the flake's side of the platform. Ultimately, the core's side of the platform will stop and actually reverse its direction. When the core's side of the platform reverses its direction, there is a rapid increase in the tearing force at the crack tip and a hinge termination results (Figure 2). So if a knapper is operating in the dynamic loaded mode, he wants flakes to exit the core's surface prior to the core's side of the platform reversing its direction. As a result, if one assumes equal platform strengths, dynamically loaded flakes are always shorter than statically loaded flakes, regardless if they terminate in a feather or a hinge.
So how does the knapper control the loading mode in which they are operating? To answer this question, let's begin with Figure 3. This is a graph of force on the platform versus time. At time equal to zero there is no force on the platform. As time passes, the force on the platform increases until the platform strength is reached and the crack initiates. The important item here is the loading time or the time it takes to reach the platform strength. Short loading times yield dynamic loading conditions and long loading times result in static loading conditions. Pressure flaking has extremely long loading times and always is static loading. As a result, it is impossible to create a hinge termination when pressure flaking.
Percussion flaking can produce either static or dynamic loading. To understand when each will occur, it is necessary to understand that all cores vibrate or ring when struck with a percussion blow. Each core has a particular vibrating frequency that is a function of its material, size, and shape. The ringing is a sinusoidal vibration as depicted in Figure 4. Frequency is expressed in cycles per second. The recipocal of the frequency is called the period and it is the length in seconds of one complete cycle. To achieve static loading conditions, the loading time (Figure 3) must be longer that the period of the core. If the loading time is less than the period of the core, then dynamic loading conditions exits.

Figure 5 presents periods for various cores, which are really bifaces supported on the edge away from the force application location. The cores range in width from 1 to 4 inches (2.54-10.16 cm). Possible thicknesses of the cores are plotted on the horizontal axis and the associated periods on the vertical axis. For example, a 3-inch wide core that is 1 inch thick has a period of 0.00106 seconds. This means that to load this core in the static mode a loading time greater than 0.00106 seconds must be achieved.

Loading time is controlled by four factors: 1) platform strength, 2) material and shape of the impactor, 3) rigidity of the support and 4) velocity of the blow. The effects can be seen in the following table.

 Fast Loading ConditionsSlow Loading Conditions
Platform Strengthweakstrong
Material and Shape of Impactorhard, high density, spherical shapedsoft, low density, long & narrow
Rigidity of Supportrigidloose
Velocity of Blowfastslow

To illustrate how all this works together, let's assume I am percussion knapping a 3 by 3 inch (7.6 by 7.6 cm) core or chunk. Figure 5 indicates it has a period of 0.000354 seconds. Let's further assume I am using a similar sized hard hammer with a slow delivery that is able to initiate a flake with a loading time of 0.001 seconds (Figure 3). As a result, I am knapping in the static mode because the loading time is longer than the period of the core. As I reduce the core to 2 inches (5.1 cm) thick its period increases to 0.000532 seconds, which is still in the range of the static loading. As the core is further reduced to 1 inch (2.5 cm) thick, the period of the core increases to 0.00106 seconds and I am now knapping in the dynamic mode. When this happens, a short flake with an ugly hinge termination occurs. And I ask myself, "what the hell happened"? I might try to slow the delivery of the impactor a little, but since I am an accomplished knapper, (not really), I have already been using the slowest blow I can and still achieve the platform strength. The same is true for the support; I have been holding the core very loose. So, to return to the static loading mode, two options remaining for me are to build stronger platforms or change the impactor. Changing impactors is much easier than building stronger platforms, which is time consuming. I can choose a larger hard rock or an antler billet, which is smaller and more manageable. Since the core is getting smaller, I choose the antler billet, which has a longer loading time because of its shape and elastic material, and return to knapping in the static mode.

The above scenario can also be visualized in Figure 6. When I began to reduce the core, which was 3 inches (7.6 cm) thick, I was operating with a loading rate to the far right of the period of the core. As I thinned the core, the period of the core increased and the period line in Figure 6 moved to the right and toward my loading rate. Finally when the core was reduced to 1 inch (2.54 cm) thick, the period line moved past the loading rate and I transitioned from the static to the dynamic loading mode.

In the archaeological record, "hard hammer" percussion was the technique used by most people at the quarries. By using slow impactor velocities and loose supports, static loading conditions were achieved in the early reduction. Referring back to Figure 5, cores with width-to-thickness (w/t) ratios in the range of 2 have little increase in periods with thinning. However, by the time the w/t ratio reaches 4, there is a significant increase in the period with additional thinning. Over the years, I have observed that most abandoned cores (bifaces) at quarries have a w/t ratio in the neighborhood of 4. I suspect, at that ratio, the knappers were having difficulty remaining in the static-loading mode without switching to a "soft hammer". "Soft hammer" is a high maintenance impactor. Therefore, if the objective is to only obtain tool flakes at the quarry it is easier to abandon the core with a w/t ratio of 4 and start over with a new, thicker one that will permit the continued use of the "hard hammer". As a result, the quarry sites contain bifaces that rarely exceed w/t ratios of 4.

Modern knappers have introduced a new strategy to the art of knapping. Instead of operating in the static loading mode, the majority of modern knappers choose to operate in the dynamic mode. To explain this logic I will refer back to Figure 6. Flakes created in the dynamic loading mode that are made with loading rates in the neighborhood of 95% of the period are extremely short and terminate in a hinge. This is because the core's side of the platform, which is moving in the same direction as the flake's side of the platform, has very little velocity and will quickly stop and reverse its direction. Flakes made with loading rates significantly less than the period are much longer because the core's side of the platforms have much higher velocity at the time of platform separation. As a result, it takes much longer for the core's side of the platform to reverse its direction. During this extended time, the crack can travel a significant distance before the hinge termination. To eliminate the ultimate hinge termination, the modern knapper selects an angle of blow that will cause the flake to gradually move toward the surface of the core and terminate in a feather prior to the hinge occurring. This gives the flake a slight wedge shape with regards to thickness with the thickest part at the platform end. To achieve the short loading times necessary for this strategy, the modern knapper uses lead impactors with a surface of copper. In addition, high impactor velocities and extremely rigid supports are selected. With this strategy, it is very difficult to make a crack travel the entire width of the core. However, flakes that reach past the midline are common.

In summary, there are two loading modes, static and dynamic. The static mode can be achieved with either pressure or percussion and the flakes never terminate in a hinge. The vast majority of flakes in the archaeological record were made in the static mode because it yields the longest flakes. The dynamic mode can only be achieved with rapid percussion loading and will terminate in a hinge if the trajectory of the crack is not forced to feather out first. The modern knapper that is employing a copper encased lead impactor is operating in the dynamic mode. The knapping techniques of the two modes are different and, therefore, when an antler knapper is discussing technique with a copper knapper they will argue about the velocity of the impactor and the rigidity of the support. Each believes the other is wrong but, in fact, both are correct.



Return to Paleoindian & Other Archaeological Stuff