# How do I measure the time taken for a laser to cut a metal?

I am a placement student and was given a project asking to measure the time taken for the laser to cut a metal and send it to Excel into a spreadsheet automatically. I am a computer science student and am unaware of how I should begin this project or how to do it. How do I measure the time taken for the laser to cut each metal?

• Given your comment to the answer, have you checked to see if there is some sort of log created by the laser’s computer that you could access? It may even be able to send a csv file after each cut or operation... – Solar Mike Jul 3 '19 at 19:11
• Don‘t forget to vote and accept the answer, which answers your questions, welcome to engineering ! – morbo Dec 30 '19 at 17:09

Cut various metals of various thicknesses, across various lengths, with the laser in question. Measure the time taken, using a digital stopwatch. Create a model that reproduces these observations, using the fewest assumptions and the simplest form, and accounts for the different type of metal, and different thicknesses. Apply the model.

You'll find such a model built into most laser-cutting software interfaces: the software will typically give a fairly good estimate of how long a given cut will take, once it has the pattern and has been told what type of material is to be cut.

• Thank you for the answer.We want to take the time over like 10 hrs. Its not possible to use stopwatch for each of them. Is there a way to make or build something simple that measures the time taken and that can be linked to excel. The laser i will be working with is based in the factory, will it also have the above said software to take time or if there isn't do u know i could input one myself? – anisha Jul 3 '19 at 9:35
• @anisha: Why can't you measure the cutting rate for a sample piece and then calculate based on the total length of cut for each piece. Your question is missing details on the name and version of software used to control the cutter so it is impossible for anyone to say what is possible with it. – Transistor Nov 30 '19 at 15:06

To accurately measure the time of a system is no causal task. You cannot taken samples and extrapolate since you will be ignoring far too many variables such as

1. material thickness
2. linear acceleration/deceleration
3. Jerk
4. Material heat absorption
5. Curve accel/decel

The best way to do this would be to read generated gcode and calculate your time with a script that will read the NC or GCODE that was generated and as pointed out in comments, getting machine specific settings, this benefits the fact that the operator who knows the material best will knowingly or subconsciously choose the correct parameters and gcode settings (which will adjust for workpiece specific properties and you will read the machine time ahead of cutting.

The machine will follow this code as accurately as it can and your script will calculate the time based on that with minimal error.

After creating a code that is accurate to a minute or seconds, can you then optimize visa experimenting and interpolation.

Have a look at the prusa slic3r github for a program that does the same for 3d printing. Which as of 30.12.19, can quite accurately predict printing time.

If your system is more ‚manual‘ and/or half automated, you can create a electronic (a voltsge divider for example) that gives you a digital high/low every time the laser is on/off, and you simply write a timer that runs. After taking hundreds of samples one could write script that predicts cutting time , but it won‘t be particularly accurate as soon as you start producing something against the ‚norm‘.

• Calculating the process time from the gcode will miss your 2nd, 3rd and 5th example of unknown variables, since they don't appear in gcode. They are "deeper" in the CNC than you could see from gcode. Slicers for 3D-printers do exactly the same and always underestimate the process time due to that reason. – OpticalResonator Dec 30 '19 at 18:38
• One will have to do some math and find some settings from the machine, this is only a matter of reading config files, i do agreee, however your misinformed slicers generally are accurate within the minute nowadays, the linked slicer given is a good example. – morbo Dec 30 '19 at 18:41
• Oh that's nice, I didn't know that, I am using a different slicer. Thanks for the info! – OpticalResonator Dec 30 '19 at 18:52