Calculating PLA packing density

[whpthomas]

So I just wanted to get someone to confirm the math.

The makerware default skinforge profile has a filament diameter of 1.82 and a packing density of 0.85. While this appears to work quite well with makerbot 1.75mm filament, I have a bunch of filament that is closer to 1.96 ~ 1.70mm so I would rather have a packing density that calibrated to my filament diameter so I can make meaningful adjustments to account for different rolls.

I while ago I ran a bunch of calibration cubes using this 0.85 packing density and found that setting the filament diameter to 1.90mm was a better setting for my 1.69mm filament. So I basically want to convert 1.90/0.85 to 1.69/???

So I am assuming I get the area A for the filament where

A = PI x (R ^ 2) 

A = PI x ((1.9 / 2) ^ 2)

A = 2.835

Then the corrected feed F for packing density P is

F = A * P

F = 2.835 * 0.85

F = 2.4

Then the area A2 for my actual filament with a diameter of 1.69 is

A2 = PI x (R ^ 2) 

A2 = PI x ((1.69 / 2) ^ 2)

A2 = 2.243

So finally the correct PLA packing density P2 is

P2 = A2 / F

P2 = 2.243 / 2.4

P2 = 0.93

I had been using 0.97 so I am going 0.93 a try and report bad on how it goes.

[whpthomas]

So far, this 0.93 packing density seems to be giving similar results to 0.85/1.90

[hellphish]

I get the best results from setting packing density to 1.0 and entering the actual diameter of my filament.

On Fri, Jan 11, 2013 at 5:31 AM, whpthomas <m...@henri.net> wrote:

So far, this 0.93 packing density seems to be giving similar results to 0.85/1.90

--
 
 

[whpthomas]

I get the best results from setting packing density to 1.0 and entering the actual diameter of my filament.

Like I said, I have tried 1.0 and 0.97. These calculations were based on running and measuring numerous calibration cubes.

What I seem to be finding is that different filament colours have different packing densities.

Clear & White = 0.97 ~ 1.0

Translucent Orange, Green, Blue, Red, Yellow = 0.93 ~ 0.97

Black & Silver = 0.93

Maybe the pigment has an effect.

[whpthomas]

The thing I was actually wondering was whether the packing density applies to the area/volume calculation or just the diameter - anyone know? The would effect the conversion from 1.90/0.85 to 1.69/???

[Susan Shelley]

I know this is an old thread now, but I've been looking onto packing density and this is to do with the hardness of the material. 

As the material goes past the driver, softer material (ABS) can get squashed and therefore has a smaller diameter when it reaches the chamber/nozzle, hence the packing density of around 8.5 for ABS. PLA is a harder material and will get squished, and maintain a diameter closer to the one measured. Therefore it has a higher packing density closer say 0.9-0.97. 

[Dan Newman]

You're on the right track. But it's not the diameter of the filament
which matters, it's the volume per extruder step or pinch gear revolution
which matters. The volume of the filament doesn't really change: the
filament is squished and is no longer circular but the area of its cross
section is unchanged as the material is effectively incompressible.
(If the area was decreasing, then you'd have a growing bulge at the
pinch gear -- the volume of material "squeezed" out -- which would very
quickly build up and jam the system. If the area was increasing, then
you'd quickly stretch the filament so thin that it would break.)

But what does change with deformation is the effective radius of the
pinch-gear/filament system. As the pinch gear turns 360 degrees it
would move 2*pi*radius linear units of filament and the volume then
would be area of the cross section of the filament times that linear distance.
Since the area of the cross section is unchanged -- incompressible --
the question then is what is the effective radius we need to use in
the calculation? This radius is smaller for ABS because the pinch gea
bites into the softer ABS filament more than PLA and so
the filament rides closer to axis around which the pinch gear is
turning. Thus, for ABS less volume is fed per revolution of the
pinch gear. And, if you look at the arithmetic used by the slicers,
you'll see them (at least Skeinforge / RepG) dividing by this factor.
So, if you use 0.85 for ABS, then you end up wanting to turn the
pinch gear and additional ~18% vs something with a filament packing
density of 1.0.

Now, that's not the entire story as the filament packing density is
also being used as a bit of a fudge factor for fine calibration as well.
It's being use to account for other issues in the system as a whole.

Dan

[Bob Hamlet]

This has been on my mind since I upgraded my extruder setup.  I was wondering what the effects of a spring with too much force would be.
The 2 possibilities I considered:
More drag on the stepper motor shaft.  Not good.
More compression of the filament into the teeth, or squashing it into an oval.  Not sure.

I did think about the new geometry of the filament if the effective center line was closer to the center of the shaft.
I was pretty sure that less plastic would be extruded, but I did not know how to account for that.
The filament packing density was not even on my radar, but it makes perfect sense!

Thanks Dan.

Bob

[Jetguy]

Answer, none of the above. Trust me, I have built more extruders than most dream of. I put the biggest spring I can on every one. No, you won't smash the filament, no it won't drag on the extruder, in fact, if anything, it's the best thing you can do to ensure proper feed that NEVER slips.

Be Nike=just do it!

[Dan Newman]

> This has been on my mind since I upgraded my extruder setup. I was
> wondering what the effects of a spring with too much force would be.
> The 2 possibilities I considered:
> More drag on the stepper motor shaft. Not good.
> More compression of the filament into the teeth, or squashing it into an
> oval. Not sure.

If you're using PLA, it's darn hard stuff and hence the values of
around 0.97 which people use. It might be deforming as much from
mechanical heating as it is from the pinch gear physically deforming it.

> I did think about the new geometry of the filament if the effective center
> line was closer to the center of the shaft.
> I was pretty sure that less plastic would be extruded, but I did not know
> how to account for that.

Yup, basically a scaling factor to arrive at an effective radius. And MBI's code
uses a radius of 10.58 mm for the pinch gear scaled by the filament packing
density. If you measure the radius at the trough of the pinch gear, you find
get a number closer to 8.5 mm than 10.5 mm. (Note also that MBI, when they
sold these pinch gears, stated they were 10.56 mm, but their code used 10.58 mm.)

Dan