Hitman

[charles.yates]

Thingiverse seems to attract models that are artsy and impossible / absurdly difficult to produce with consumer-grade printers.

I'd amputate the wings & tail & maybe the head, print the pieces aligned in the obvious orientations, then glue everything back together. Might be faster & less frustrating than invoking heroic support structures ...

Yeah, I had cut mine in half for one attempt in ABS - got the upper half (wings and head) OK [slightly frayed at the edges of the wings, but not too bad]. Bottom half screwed up with tail rocking out of place. Probably should have tried cutting the tail off too - might give that a shot if jsc's model doesn't help.

[charles.yates]

I printed a Charizard. I don't think it was exactly that one. I don't remember having issues with the tail or supports, but the wing on mine was so thin it would disappear at any reasonable scale when sliced. I beefed up the wing membranes with Meshmixer (model attached).

It sounds like you may be having some bed adhesion issues. Try:

1. Using some adhesion aids on your bed. Glue stick or hairspray work very well; if you're not using either, give it a shot.
2. Calibrate your extrusion.
3. Add a small amount of extruder lift on retraction (.1 - .2mm) to ensure you are not contacting any over extruded bits. Any contact at all on a support will eventually knock them over.
4. Slic3r comes with several different support styles. Try them all!

I have a .25 lift at the moment and have tried all of the supports at some point or other (not all with this particular model though - ended up rectilnear at the time, but may dabble with that again).

I have callibrated the extrusion with the 10cm straw and the pen approach, but admittedly, it was with another filament - could try that again I guess.

Adhesion may help with the tail I guess - worth a shot - thanks for that and the modified model.

FWIW, I tried scaling the model I had down by 0.5 and slicing as I mention above, but that didn't work - suddenly had some supports missing and parts of the wings in midair . Not much point trying that then .

[charles.yates]

[quote="charles.yates"Split the model right down the center and print each half laying flat - way less support and more area to apply the support to where it is needed. Then just glue them together.

Also worth a shot - thanks for the suggestion.

[charles.yates]

Curiouser and curiouser...

This was my first attempt .

After checking some of the things which jsc advised - improving adhesion (I put down new kapton as it's normally been enough), calibrating the extruder, and using his model, I've arrived at:

Many thanks for all the tips guys .

I have to say, I'm not too interested in the model itself, but I am interested in getting the printer as good as possible, so really appreciate the help from all of you.

[jsc]

Is that... better? It's hard to tell what's going on. It does look like you're over extruding somewhat. Your support walls shouldn't have those striations and weird bulges.

[jsc]

I just looked at that calibration link. You don't want to be messing with the extruder steps. Instead of trying to measure the length of filament used, you should directly measure what you care about, the width of your actual extrusion vs. what you asked for. Try resetting the extrusion steps to the default (471.5) and follow the procedure here: viewtopic.php?f=3&t=1964

[charles.yates]

Is that... better? It's hard to tell what's going on. It does look like you're over extruding somewhat. Your support walls shouldn't have those striations and weird bulges.

Definitely better, yeah - it did complete at least , but the wings came out frayed (left and right extremities completely gone) and a lot of the misplaced filament ended up collected around the mouth . I left it running when I went to bed, so I didn't see precisely what happened.

Interesting that you would suggest using the cube approach instead of the E value. Read a few discussions on this one, and well, I guess I've been more convinced by the arguments for modifying the E (though of course, my own attempt hasn't come out completely right here, it's still a massive step forward from my earlier attempt, which was, in turn, better than the default).

Having said that, I just tried a quick cube (with my E value and 20x20x2 solid as per Jules' document) - the height was 2.02mm, so yeah - over extruding seems right (which may actually have correlated with my verification run through of the E - there was a small discrepancy there, but being lazy, I kinda figured it was OK - it was less than 1% though).

I'll give both a try though as I'm curious as to which gives the more accurate results.

[jsc]

You can't tell anything from the height of the cube other than that your Z setting is very slightly off. Measure the walls, as described in that link.

[charles.yates]

You can't tell anything from the height of the cube other than that your Z setting is very slightly off. Measure the walls, as described in that link.

Yeah - had misrembered the height test - was just a quickie over lunch time with no time to double check .

Anyway, I did the test and it mostly fluctuates around the 0.4mm and 0.41mm reading, though there are a couple of spikes - dropping to about 0.37mm and rising to 0.46mm. The height is a fairly consistent 2mm after adjusting the Z offset in slic3r.

This is with the E calibration I did last night (494.72 was the figure I settled on), 1.75mm diameter filament and extrusion of 1 (I checked the filament again and reduced the extrusion from 1.05 to 1 before running the thin wall test - just wanted everything as 'normal' as possible).

I should also point out that slic3r has a hint tip which suggests that if you're changing the value of the extrusion outside the range of 0.9 to 1.1 you should "check filament diameter and firmware E steps". So I'm still feeling kinda justified on that one.

Prior to this, I ran a test using Bratag's suggestion of simply splitting it down the middle and printing the twos halves on their side - the model you gave me didn't survive the process, but the original model did and that worked fantastically well - sadly, I hadn't noticed that it wasn't quite oriented correctly, so the right hand side wasn't an exact mirror of the left which would make the join a bit odd and the half I printed suffered a bit in the process (also, I'm a bit anally retentive ). I will probably try this one again over the w/e and post some pics if it turns out OK.

Half print attached - 0.3mm slicing and showing (possibly over extrusion too?):

[jsc]

If you're just doing a simple plane cut, you can do it without any additional software. Just duplicate the object, rotate one 180 around x or y, and drop both below the plane of the bed to the same cut line. A little simple arithmetic may be involved.