Laser-Cut Acrylic: Fit Tolerances

February 25, 2014 Engineering

As I am now fully trained on the Trotec 500 laser cutter at NextFab Studio, I have begun experimenting with acrylic as a material for my Masters Project on web-app customized products.  Using code and user input from the web app, I would be able to automatically generate the design files necessary to create customized products, which could then be cut using a variety of rapid prototyping tools.  In order to design with any material, one must first know the limitations and associated parameters.  So, I designed a set of test coupons to determine what tolerances I should use when designing for 1/8″ sheet cast acrylic.  This material comes in a wide variety of styles including transparent, translucent, opaque, marbled, sparkly, etc, and would thus be perfect for customization purposes.

The first step was to measure the material.  Though labeled 1/8″ thick, cast acrylic sheets are rarely the exact thickness they claim to be.  So, the material must be measured to determine the actual thickness and design accordingly.  Because the sheets come backed with paper, I decided to measure the thickness of the paper, and then the overall thickness of the backing and plastic sheet, subtracting to find the actual thickness of the material.  Multiple thickness measurements should be made around perimeter of the sheet to account for variations.  I made 10 measurements at different locations around the edge of the sheet.

Measurement [mm]
Paper Thickness 0.07
Overall Thickness Measurements 3.09
3.17
3.18
3.05
3.03
3.12
3.08
3.01
3.07
3.07
Average Overall Thickness 3.087
Average Sheet Thickness 2.947

  • Measuring the backing paper.

The next step was to design the coupons.  For making structural parts, it is important how the manufacturing process will affect the end result.  In casting and injection molding, shrinkage due to rapid cooling is a major concern.  For welded assemblies, its warping due to the concentrated heat loads applied at the joints.  In laser-cutting, the laser is actually vaporizing some of the material, so the thickness of this cut becomes important for determining final fit of components.

So, I designed a set of coupons in Solidworks with an incrementally increasing slot dimension to test the mechanical fit.  To I made these into extruded solids and then saved the faces as a DXF file.  From there, I imported this file into Adobe Illustrator, where I created the mating component, and applied numbers to tell them apart.

  • Parametric design of the test coupons.

After that, I cut out the design using NextFab’s Trotec500, and here’s what I got:

A_DSC_0209

 The next step was to take detailed measurements of each sample for analysis in a spreadsheet.  I chose to measure the length and width dimensions, as well as the slot dimension.

A_DSC_0229

From here, I compiled the following spreadsheet:

Sample # Nominal Actual Difference
Length [mm] Height [mm] Slot [mm] Length [mm] Height [mm] Slot [mm] Length [mm] Height [mm] Slot [mm]
1 30.00 20.00 2.60 29.73 19.87 2.71 0.27 0.13 -0.11
1 30.00 20.00 2.60 29.74 19.83 2.85 0.26 0.17 -0.25
2 30.00 20.00 2.75 29.76 19.61 2.88 0.24 0.39 -0.13
2 30.00 20.00 2.75 29.75 19.67 3.08 0.25 0.33 -0.33
3 30.00 20.00 2.90 29.75 19.66 3.07 0.25 0.34 -0.17
3 30.00 20.00 2.90 29.80 19.63 3.09 0.20 0.37 -0.19
4 30.00 20.00 3.05 29.80 19.86 3.31 0.20 0.14 -0.26
4 30.00 20.00 3.05 29.75 19.94 3.32 0.25 0.06 -0.27
5 30.00 20.00 3.20 29.79 19.63 3.46 0.21 0.37 -0.26
5 30.00 20.00 3.20 29.77 19.69 3.50 0.23 0.31 -0.30
6 30.00 20.00 3.35 29.78 19.64 3.67 0.22 0.36 -0.32
6 30.00 20.00 3.35 29.75 19.74 3.55 0.25 0.26 -0.20
7 30.00 20.00 3.50 29.80 19.61 3.72 0.20 0.39 -0.22
7 30.00 20.00 3.50 29.78 19.66 3.89 0.22 0.34 -0.39
Average Difference [mm] 0.23 0.28 0.24
Laser cut width [mm] 0.25
Line Shift [mm] 0.13

By calculating the difference between the nominal dimension, as I designed it in Solidworks, and the actual dimension, as measured on the sample, we can find the approximate width of the laser cut line.  From this batch of samples we find that the laser cut width is approximately 0.25mm, meaning that the laser shifts the cut line about 0.13mm to eitherside of the center.  This means that in order to get a perfect 30mm x 20mm rectangle, the cut-lines should be dimensioned at 30.25mm and 20.25mm.  For a perfectly cut rectangular hole, you’d dimension the cut-lines at 29.75mm and 19.75mm.

When I put the coupon halves together, I got the following results for fit:

Sample Nominal Slot Dimension Fit
1 2.60 Did not fit – Too Tight
2 2.75 Tight Fit
3 2.90 Loose, Sliding Fit
4 3.05 Did not fit – Too Loose
5 3.20 Did not fit – Too Loose
6 3.35 Did not fit – Too Loose
7 3.50 Did not fit – Too Loose

A_DSC_0214

In the end, only two of the sets fit together.  Sample 2, where the slot was undersized by 0.30mm from the nominal 2.90mm sheet thickness had a nice, tight fit.  The parts had to be pushed together, and once together, had to be forced apart.  Sample 3, which had the nominal slot thickness of 2.90mm, was a very loose, sliding fit.  It slid together nicely, and easily slid apart with minimal play in the joint while together.

Taking these factors into account, here are my recommendations for sizing cuts in acrylic:

Fit Dimension Modifier [mm]
Loose Sliding Fit 0
Tight Sliding Fit -0.08
Tight Fit -0.15

If you’d like a loose sliding fit, just size the slot for the sheet thickness.  The cut of the laser will take out more material, leaving a small gap between the pieces, allowing easy sliding motion without too much play in the joint.  For a tight sliding fit, subtract 0.08mm from each side of the slot, for an overall undersize of 0.16mm.  For a tight, non-sliding fit, undersize the slot 0.15mm on each side, for a total of 0.30mm undersize.   This will produce a tight structural fit, capable of maintaining a rigid connection without acrylic cement.

My next experiment will be to create some snap-fit clips that could be generatively designed using a code template and some user input geometry parameters.  In researching this topic I found some other helpful guidelines for working with acrylic from FabLab@School and Deferred Procrastination, you should definitely take a look!

In retrospect, I’m not sure why I thought the slots would need to be so oversized, e.g. samples 4 through 7.  However, this experiment definitely reinforced the benefit of knowing your materials and their manufacturing processes before designing things.

This article has 3 comments

  1. Ian Johnson

    I have been going through this while working on some slice forms with 123D Make. The main challenge has been that the tolerance of the sheet thickness is .1mm or more, which can be enough to make the difference between slots that are too tight, and slots that are too loose to stay connected while you assemble the object. I’m starting to find that much like measuring your filament before starting a 3D print, I have to measure the sheet that I’m going to use and try to find a section that doesn’t vary too much. Then I generate the drawings in 123D Make with a material thickness that matches the sheet (or portion of a sheet) I picked out.

    There seems to be an acceptable range of sizes between too loose to hold and tight enough to risk breaking. The trick is to find the min and max thicknesses and size the slot for the median, and hope the thickness doesn’t go out of bounds for any of the slots.

    Something worth looking into is if there is anything that can be done with the shape of the slot to maximize the range of material thickness that will work with it.

Leave a Reply

Your email address will not be published. Required fields are marked *