The printer was delivered crated up on a pallet truck - it is heavy and required 4 men to lift it onto the trolley we had made for it. It was commissioned by Alex Wood (Techsoft) who also provided training in how to use it.  This only took two hours.

 

The very first job to be printed (by Alex) was our keyfob torch case. (This project was conceived to provide pupils with an introduction to proDesktop.  proDesktop designed cases were 3D printed by Stuart Douglass at Ripley St Thomas on their SLA printer.  The cases cost £1.30 each and were funded by Rolls-Royce.  The proDesktop worksheets can be found here.)

 

 

Design rendered in proDesktop

 

 

 

SLA Print of original key fob torch case

 

 

The resulting model was impressive, although compared with the SLA printed ones the circular shape had acquired 'flats'.  Clearly, this printer was capable of making our torch cases without us have to send the job else where for printing.

 

 

Our first print job on the Dimension printer

 

 

The school's open evening was due to take place a few days later.  Obviously, the printer was worth putting on show.  We switched on the printer and prepared to load a job but we could not connect to it.  We could not even ping it.  Then we remembered; we had in the interim restructured the department network which meant changing the net mask of all the computers.  We had forgotten to change the printer's setting.  Changing the setting we could successfully ping the printer.  We launch CatalystEX, delete its now incorrect printer (because we have changed its address) and add a new one (with the correct address).  We load the torch case, process it and send it to the printer - but it does not arrive!  The file seems to disappear into hyperspace.  It turns out that in replacing the printer driver with a new one with the corrected address, we had not selected the correct Dimension!

 

 

Strangely, the excellent and sophisticated CatalystEX does not appear to give an error message telling you that the printer at the designated ip address is not the correct model given in the 3D printers list.  Still, once corrected the file is transmitted and received by the printer. 

 

Learning point:

 

 

We press the button to start the build.  The mighty machine launches into its sequence of actions to build the part.  It looks for home, it asks if we have seen it purge . . . . we haven't so we reply no. It tries to purge - still we do not see it purge. After several attempts to purge we decide to read the detail in manual.

 

The manual gives several scenarios associated with failed purging.  In the end, we go for the most involved remedy.  We swing the head forward, cut the filaments and remove both cassettes pulling the filaments clear of the machine.  We unbolt the feed motors (as described in the manual) and remove all traces of filament.  We clear plastic from the liquefier tubes and use the small allen key to clear them.  We close the print head, install the cassettes and set the machine to load.  We hear the filament unwind and then engage with the pinch rollers, but still no purging.  We clear the head a second time, this time we do not unload the cassette but cut the filament at the head and clear the pinch rollers.  We then switch on the head motors (all as described in the manual), pull the filament from its exit and manually feed it into the pinch rollers - it purges beautifully.  The motors are run as we close the head.

 

We again start to build the model.  A few minutes into the build, when the printer tries to lay down the base we see no plastic being extruded.  A phone call to Alex; he suggests we remove the cassettes, clear the head and reload the machine with new cassettes.  We do all this, we hear  the filaments load and engage, but still no purging!  Time to call someone in.

 

The service engineer appears only two hours later.  He carries out the same tasks that we had done.  He checks that material purges and runs the bottle test part!! Here it is . . .

 

 

Sample bottle test piece (the code for this resides in the machine)

 

. . . . clearly we do not have the right touch!

 

Close questioning of the engineer reveals two learning points.

 

 

 

 

The sample bottle shows what the machine can do.  For a start, it prints the lid and the bottle side by side.  After breaking away the support material the lid can be screwed onto the bottle.  Secondly, the wall thickness of the bottle is only about 1mm.  Perhaps we can reduce the thickness of the torch case, from 2mm to 1mm and halve the cost (and shorten the build time). The new design includes four pips to retain the base.

 

 

Redesigned keyfob torch case

 

 

Before trying to print the new design we looked at why the first print had pronounced flats.  The cause was an inappropriate chord length when exporting from proDesktop to stl file.

 

 

Settings used for outputting to stl file

 

 

We exported to a new stl file with a chordal length of 0.01mm and angular deg of 1.  In CatalystEX after processing it looks like this:

 

 

 

 

The problem now was that we could not add the design to the pack.  It also looked different, why were some slices coloured yellow?  Eventually the problem was revealed  by a message at the bottom of the window . . .

 

 

. . . the error message is rather inconspicuous.  Now we know where it is we can keep an eye on this area of the screen.

 

Here are another couple of learning points.

 

 

 

 

Reimporting the stl file into CatalystEX and looking careful at the rendering shows where the problems are.

 

 

 

 

Rounding of the key ring loop cannot be done where it changes direction at the main body of the case.  (The material thickness set at 1mm had a 0.5mm rounding). This is rather fine detailing for the machine drawing with a 0.2mm thread of ABS.

 

Rounding of the loop was removed.

 

 

 

 

The model was turned over before slicing to reduce the required amount of support material.

 

 

Slicing was now successful - no yellow lines.  Model material is shown in red, support material in white.

 

 

 

The build produced a superb torch case which still has strength but is much closer to retail standards.

 

 

The modified torch case rendered in proDesktop

 

 

 

The printed torch case alongside its internal electronics

 

 

 

The complete assembly with back made from 1mm HIPS

 

 

 

How the torch looks from the front

 

The back of the torch is laser cut 1mm HIPS.  It includes four semicircular cut-outs to allow it to fit past the four retaining pips.  The back is locked into place by rotating it using the two holes.  This is a little tricky as the back is quite a tight fit.  What we need to do is make a key that locks into the two holes to give a good purchase. Ah . . . I know just the machine for doing this.

 

The modified 3Dprinted case uses approx 60p worth of material and takes 20 minutes to build.

 

 

Video showing the torch case build