update on air-proofing ABS prints

We are still struggling with the porous 3D prints. However, we wanted to keep you updated on the progress that is happening:

We changed to ABS, since acetone vapour can be used to finish the surface of the individual parts. A “vapour chamber” with a “vapour circulator” has been constructed. Below some snapshots:

The spot light is one option to heat the acetone camber. A radiator was used successfully as well. The aim is to create a saturated acetone vapour inside the chamber. There is no need to boil the acetone.

Below you will find images, that show you the results before and after the finish:

Layer thickness of 160um offers the best results so far. But, as you can see, the parts are still not air-proof.

Issues and solutions:

  • There are problems with the slicer software. A non-solid layer is included into the G-code. A new slicer software should solve this problem
  • Moisture inside the ABS filament would result in a more porous print !? Thus, we are planing to pre-heat the filament, in order to remove the moisture.
  • The vapour smoothed parts look quite good, once they are removed from the chamber. But after drying the cracked finish (as in the photos) appears. Here we are planing to avoid a change in temperature during the drying process.

choosing the optics for the resonator

The laser will be comprised of more or less cheap and easy sourceable components. This is an important requirement in order to make this development available to the maker community.

The only possible issue are the optics for the resonator. There are suggestions (see Sam’s FAQ) to use polished pieces of copper as a mirror and to drill a hole into the copper as an output coupler. However, using such approach it seems difficult to achieve high power outputs, as desired by the laser. Thus, specialized mirrors and output couplers need to be purchased.

A cheap (and in my research, the only) option is to rely on an import from China. On Ebay there some companies offering CO2 laser optics. It is important to look for resonator optics, i.e. output couplers and concave mirrors.

Thus, the discussion of the resonator design follows the available resonator optics.

  • output couplers: On Ebay, there seems to be two different kind of material output coupler: germanium (Ge) or zinc selenite (ZnSe). Both materials are coated to obtain the offered reflectivity of about 70%. The output couplers are plane. One advantage of ZnSe compared to Ge is its transparency in the visible. A laser pointer can be used to align the resonator.
  • mirrors: On Ebay available mirrors are concave with an curvature radius varying from 2m up to 10m. The substrate is silicon or glass, that is coated by gold or/and additionally dielectric coating. One should prefer to use the silicon mirrors for higher power application. Since, the coatings achieve no 100% reflectivity and, hence, some of the laser power is transmitted to the substrate. Glass absorbs at wavelength of 10.6 um and the glass mirrors may be damaged.The requirements on the curvature are discussed below.


The resonator geometry is spherical-plane, as dictated by the available components. Then, the resonator needs to be stable, in order for a laser mode to form within the laser. If the resonator is unstable, the losses within the laser might be to large and no lasing is achieved.

In order to obtain a stable spherical-plane resonator, the distance L between the mirror and the output couple has to smaller than the curvature radius R of the mirror, i.e. L<R.

This requirement is easily meet for our 50 cm long glass tubes separating the mirror and output couplers. A mirror with R=2m would form a stable resonator. But, what are the advantages, if there are any, by choosing larger radii?

The crucial parameter to be considered is the beam radius. It is important to note that the laser beam is by no means cylindrical. The beam diameter varies within the resonator and outside, as shown in the following figure. The beam is indicated in red and the outputcoupler and mirror in blue. In this case, the laser would propagate to the left.

beamOutside 0.5m -2mFurthermore, the intensity of the beam is the highest at its centre and drops exponentially off to its sides. (At leased, this is the case for the simplest mode, the fundamental mode.)

The calculation of the mode properties are done in Mathematica (see pdf ). The pdf contains all formulas and references to literature. Some of the possible configurations that are possible with the available components are highlighted below. (left blue line: output coupler, right blue line: curved mirror with radius R, thick red lines: beam contour containing 86% of the laser power, thin red lines: beam contour containing 98% of the laser power, dotted lines: beam diameter on mirrors, dash-dotted line: inner bore diameter of discharge tube)

  • L=0.5m, R=2mResonator 0.5m 2m
  • L=1m, R=2m
    resonator 1m 2m
  • L=0.5m, R=5m
    resonator 0.5m 5m
  • L=1m, R=5mresonator 1m 5m

There is a trade-off between beam divergence, beam diameter, discharge volume and diffraction losses. For instance, in the case of L=0.5m, R=2m, the mode is well contained within the discharge tube. The diffraction loss will be small. However, the gas discharge volume is larger than the mode volume. Thus the gain is not optimum.
The other extreme is the case L=1m, R=5m. The 98% contour line does not fit in the discharge tube, thus the diffraction losses are larger than 2%. This limits the output power of the laser, as well.


We will use a ZnSe outputcoupler (70% reflectance) and a dielectric coated silicon substrate mirror with an radius of R=2m.

good and bad news

good news: there has been some progress on the groove side. We managed to figure out a design for the groove that allows us to print the overhang without any problem. Also, we are now able to push the discharge tube into the end caps for a groove depth of 2.2 mm (P5) and even 2.0mm. For the shallow groove  an applicator was printed that spreads the o-rings as the glass tube is pushed into the end cap.

bad news: the printed end caps are still not airtight. We have been experimenting with a concentric layering, 4 layers wall thickness and  a filling fraction of 80%. But no luck so far. The next step is to try a solid (100%) print.

Since we are using PLA for printing, it is somewhat difficult to use solvent vapours to smooth (and seal) the surface. According to wikipedia, “PLA is soluble in chlorinated solvents, hot benzene, tetrahydrofuran, and dioxane“, which are in turn all somehow carcinogenic. There is one report, that uses tetrahydrofuran. However, due to the hazardous solvents, this seems not to be viable road without any access to a proper chemical lab.

ABS on the other hand is solvable in acetone. There are numerous account of using boiling acetone to create acetone vapour for smoothing the surface. Here is a link to you tube video that uses this technique to air proof ABS parts.

And finally, here is a time lapse clip of an end cap being printed on an RepRap Ormerod.  The entire print took about half an hour.

not really air-proof …

This is the first attempt to print end caps to the gas discharge tube.

Different grove depths for the O-rings have been used:

P1 groove depth  = 2mm
P2 groove depth =  2.4mm
P3 groove depth =  2.8mm
Diameter of O-Ring = 13.5 x 3mm


  • the 3d printed part itself is not air proof (due non-solid fill)
  • P2 + P3, the tube can be installed, P1 still to tight
  • printing artefacts (filaments threads at the to end of the groove) => too steep overhang?
  • non concentric printing path

vaccum pump arrived

The new vacuum pump has arrived. Since we are on a budget, the cheapest option from ebay has been chosen. The Model VP115 is a rotary vane vacuum pump with a pump rate of 42l/min and a final vacuum of 5Pa = 0.05mbar (according to the manual).  To be honest I am a little bit sceptical regarding those values.

However, since only a vacuum in the region of a few tens mbar is needed for igniting the gas the pump should do fine. And more importantly I found a project in Switzerland that used exactly this pump to do gas discharge experiments.

One (of the many) remaining problems is to measure the pressure of the vacuum CHEAPLY. This is important since the lasing will depend on the pressure of the gas mixture. The best option so far seems to used a selfmade Pirani gage. There a few project on the web, like for instance this or that.  But the problem is the calibration of the gage. This might only be possible using an additional calibrated gage ….

A final thought,since we are dealing with a gas mixture, that may change in its relative composition, a Pirani gage might not work at all. The heat conduction of helium very high compared to CO2 and nitrogen. Thus a small change in the composition might void the calibration…. You are welcome to comment, if you have any ideas!

first 3d printed parts

The first attempt to print the end cap:

It turned out that it took rather long to print the entire part (>7h). Due to a software problem the end cap was printed in two separate parts.

The first results are encouraging in term of the resolution of the print. An accuracy of about 0.1 mm could be confirmed and the glass tube fit nicely into the end caps.

The focus for the next steps is to redesign the grooves for the O-rings since the glass tube could not be pushed past the fitted O-ring.

long term power output of laser tubes from china

I found an interesting website that looks into the long term power output of the glass laser tubes offered by Coletech ltd via Ebay from china. These tubes are also used in cheap laser cutters that have been available for a about 600 Euros, as found in our maker space.

The laser tube is sealed by glueing the two resonator mirror onto the glass tube itself. A drop to about 80% of the power output over a time of 7 month is a relative good result for a “soft sealed” laser. Note, that the results refer to 11W, only about 25% of the available power. Its good to know that “clearance” runs can increase the output power.

Unfortunately there are no information on the degradation of the tube with working hours or output power.

the glas tubes have arrived

The glass tubes are made from borosilicate glass and were purchased from a glass blower (Eich) .  Length 500mm, diameter 46mm and 14mm