Monday, July 17, 2006


The RepRap mints a polo

Adrian has improved the infill code to the point where it can now fill in the gaps in an annular object - a short, fat tube. The shape is commonly referred to in the UK as being like a Rowntree's Polo Mint ("The Mint With The Hole"), which are now sadly owned by that ethically dubious company, Nestle.

Meanwhile, back off the soap box, I've grabbed a photo of the printed polo with all the little bits of sprue snipped off. Note that about 10 o'clock there is the dreaded dimple caused by the nozzle not having any polymer left in it after it has been moving around. Really must fix that some month soon:

We have even layers, like an onion or possibly parfait. These go round corners and still manage to line up quite nicely if I say so myself. The infill can be tightened up, but I like to see where it's going. The object was cooled for 10s by a fan between layers.

The polo was meant to be 20mm in diameter. 3 random measurements give 19.5, 19.7 and 19.6mm. The centre hole was meant to be 10mm diameter and that gives (avoiding the dimple) 8.3, 9.6 and 9.1 - there's a lot of sprue so it's hard to take accurate measurements inside.

Warning: RepRap'd artificial sweets are a choking hazzard. Keep out of reach of young children and intoxicated students.

Now THAT belongs on the front page. :-D
Looking at your measurements... two things come to mind...

#1) why is the center hole so off? from 8.3 to 9.6? and why is the hole smaller? (shrinking should have enlarged the hole, no?) How was the item cooled? was the fan overhead? or coming at it from the side?... if from the side, what side in relation to the measurements... I would be interested in seeing what would happen if you let the prototype dry from ambiant temperature, maybe putting down a layer every hour or couple hours or twice a day or something...

#2) what is causing the discrepency in sizes? I'm guessing there's a bad algorithm or something... I would expect the two diameters to be proportionally off, in different directions, if it were simply 'size' or 'shrinkage'...

I would guess the outside diameter is just a slight miscalculation on shrinkage (2.5% too small)... the inside diameter, on the otherhand, should have grown in size... and by about as much... (10.3 or so)... It kinda looks to me like maybe the 'border' on the inside of the hole is 'on the wrong side' if you know what i mean... i would guess that the outside diameter of the piece of extrusion that forms the edge of the inside circle is in that 10.3 range...
#1 The centre hole isn't actually off-centre, I checked with callipers. There is some distortion at the dreaded dimple, which gives this impression.

#2 The discrepancies are roughly 0.5mm, which is half the width of the extruded filament. This may not be coincidence.

Fear not, there is still much investigation and tweaking to come. A narrower extruder head is in the works (see previous blog entry), which may shed more light on the issue.

Vik :v)
Woah. What was the build time?
About 8 minutes, I think. I didn't time it.

Simon, could you put a timer in the deposition complete dialogue?

Vik :v)
[lord cat]The cooling was done with a fan mounted at the side of the deposition head, pointing directly down and shrouded to avoid cooling the heater. I've got to grab another picture soon, so stay tuned.

No drying is necessary. We're depositing molten polymer and it is hardened by the air blast from the fan between layers.

Vik :v)
You guys are on a roll these days!

Just an idea about the shrinking polo. If extrusion is done in a circle it might seem to shrink because the extruded material will always pack up on the inside edge and stretch on the outside rim. There is just less area on the inside than there is on the outside, so the extra has to be pushed somewhere... In this case, further to the inside...
Shrinkage is unlikely to be a major issue at this point, as the outside is deposited first as a very narrow stream. The thin layers are stuck to a rigid base to prevent shrinkage and curl-up.

Vik :v)
I've done mold melts with both HDPE and CAPA. While there is significant though not huge shrinkage in HDPE but it is hardly noticeable in CAPA.
Vik and Plaas, I'm following you guys on the shrinking issue. I guess I expressed myself poorly. I'll try again using a different word. ;)

I suspect that extruding in a circle smears the CAPA inward. You see this effect very nicely in the photo uploaded a few days ago where Vik had a rectangular outer shape with an complex inner pattern. The outer rectangle was very straight and consistent (nice work!) but at the corners, you could see that the CAPA was smearing inwards a bit, because there is less surface area on the inside edge of a corner then there is on the outer edge (aka overlap?).

Circles of a smaller diameter, like the polo, would also show this effect, but evenly distributed along the inner and outer rings. I suspect that "the anomaly at 10 o'clock"(sounds exciting, sort of like a UFO sighting!) might be amplified by the above mentioned smearing inside curves as the extrusion head closes it's pattern.

Of course, not being a first hand witness, I have no credibility when it comes to UFOs, extrusion smears, or polo mints. ;)
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]