Why Laser Cutting Speeds can be Misleading
So you are in the market for a laser cutting machine, and you want to know how fast you'll be able to cut parts. Easy enough, right? Every laser manufacturer will tell you right away how fast their machines will cut different materials. Take the cut line length for your part and divide it by the cutting speed, and you have a good guess. Right?
Unfortunately, it's not that simple. As you almost certainly know, actual laser cut processing time for a part involves many more factors, including the following:

Number of pierces,

Traverse time between contours,

Leadin length and speed, and

Machine acceleration.
We knew this when we got started, at least in theory. But we were still surprised by how much higher laser cutting time can be than what is predicted when using the maximum cut speed alone. Using line length and maximum cutting speed alone, estimated cut times can be off by more than 20x the actual cut time!
The exact numbers will vary by machine and material thickness, but we found that on our 3 kW Trumpf 1030 fiber laser  a stateoftheart machine, purchased new in 2018  the primary factor affecting cutting speed is machine acceleration. It doesn't matter if your machine can cut sheet metal at 1200 inches per minute if it can't accelerate fast enough to follow the contours on your parts at that high speed.
Cutting speed is affected more on thin materials, where the theoretical laser cutting speed is very high. Thicker materials are affected less.
To quantify the impact of machine acceleration on laser cutting time, we designed a metric to quantify the curvature, or "curviness," of the contours in a part. Curvature is measured in radians per inch, and basically describes how quickly the cut direction changes per unit length. High curvature parts require higher acceleration to maintain the same cutting speed.
We created a collection of nine similar lasercut test parts of comparable size, with varying amounts of curvature per part. Each part also consisted of a single contour, minimizing the impact of traverse time and pierce time.
Examples of the test part designs are shown below:
Curvature: 7.67 rad/inch,
Max cutting speed: 1200 in/min
Naive cut time estimate: 1.52 seconds
Actual cut time: 14.4 seconds
Cut time increase: 9.5 x predicted
Curvature: 15.32 rad/inch,
Max cutting speed: 1200 in/min
Naive cut time estimate: 1.52 seconds
Actual cut time: 23 seconds
Cut time increase: 15.1 x predicted
Curvature: 30.86 rad/inch,
Max cutting speed: 1200 in/min
Naive cut time estimate: 1.54 seconds
Actual cut time: 30 seconds
Cut time increase: 19.5 x predicted
In the figures above, the maximum lasercutting speed is the published value, from Trumpf, for 0.04" steel. These numbers are not only published for our specific machine (the 3 kW 1030 fiber laser), they are defined in the machine "tech table" as the target cut speed. For very long straight lines, the machine does actually cut that fast.
But how fast does it cut our "curvy" parts? Using the maximum cutting speed for this material thickness as an estimate, the projected cut time for the far right part is 1.54 seconds. The actual cut time for the part is 30 seconds. That is, for a curvature 30 part, the maximum cutting speed underestimates cutting time by a factor of 19.5.
It's also worth noting that this test only characterized the impact of curvature. Parts with a high number of piercings yield cut times that are even higher.
As mentioned previously, the change in cut time is most pronounced for thin materials, where cutting speeds are high. As the maximum cutting speed decreases, the impact of curvature also decreases, and the actual cut time more closely resembles the predicted.
The figure below shows how the cutting time changes as a function of part curvature, for maximum cutting speeds ranging from 47 inches / minute (3/4" thick steel on our machine), to 1160 inches / minute (0.04" thick steel).
Increases in cutting time relative to predicted based on maximum cut speed and line length, as a function of curvature. The higher the part curvature, the larger the impact. Changes in cut time are shown for materials ranging from 0.04" thick to 0.75" thick.
As shown, when the maximum cutting speed is above 1000 in/min, a contour with curvature of 30 requires over 16x the cutting time predicted using line length and maximum cutting speed alone. At a much lower maximum cutting speed of 47 in/min (i.e. for very thick materials), there is very little increase in cutting time.
In general, a higherpower laser will yield improved cutting speeds. But acceleration matters as well. Even knowing the acceleration limitations of a machine won't necessarily tell you how fast it will cut in practice, either. Perhaps the best method for determining actual cut speeds is to obtain a copy of the manufacturers nesting and path planning software  BOOST, in the case of Trumpf  and generate programs for a wide variety of parts. Seeing the predicted laser cutting time from the actual software will give you a good feel for how quickly you'll be able to cut using a particular machine.