Support material z-distance: wrong variable used #3376

Ricky1975 · 2016-06-25T13:45:51Z

Version

1.3.0-dev
1.2.9-265-gcb72967

Operating system type + version

Win7x64 (any)

Behavior

Distance between support material and first object layer is calculated wrong: $nozzle_diameter is added but not usefull, if distance is set manually. Maybe I'm wrong, but please check this.

Current SupportMaterial.pm:

sub contact_distance {
    my ($self, $layer_height, $nozzle_diameter) = @_;

    my $extra = $self->object_config->support_material_contact_distance;
    if ($extra == 0) {
        return $layer_height;
    } else {
        return $nozzle_diameter + $extra;
    }
}

==> Maybe should be:

sub contact_distance {
    my ($self, $layer_height, $nozzle_diameter) = @_;

    my $extra = $self->object_config->support_material_contact_distance;
    if ($extra == 0) {
        return $layer_height;
    } else {
        return $layer_height + $extra;
    }
}

br
Hendrik

Ricky1975 · 2016-06-25T18:29:49Z

Also a question: Why is the nozzle width always used to calculate support layer heights? Why is there a hard coded factor of 0.75?
Wouldn't it be much more efficient, if there is a user controllable parameters to define "Print Support every x layers" affecting only the layers below the interface?

lordofhyphens · 2016-06-26T00:17:19Z

Because of a general rule of thumb that you get poor adhesion/extrusion
with several layers > 75% of nozzle width.

@Ricky1975 Slic3r picks the object's layer height or 75% of the nozzle width, whichever is greater.
https://github.com/alexrj/Slic3r/blob/master/lib/Slic3r/Print/SupportMaterial.pm#L340

References:
https://ultimaker.com/en/community/2099-max-layer-thickness
http://reprap.org/wiki/Triffid_Hunter's_Calibration_Guide

lordofhyphens · 2016-06-26T04:50:25Z

Yeah, I'm not sure why it's doing this. Easy enough to test to see if it does anything weird though with the change :) Suggest running a few print tests with the variable changed (if you can run a build, it's trivial to change stuff like this and try again because you just need to restart Slic3r and not rebuild).

If anything the way it is biases at being slightly more distant from the previous layer. The distance used to be nozzle width * 1.5.

@alexrj any reason why this is the way it is? I see it's been there since the option was added.

Ricky1975 · 2016-06-26T09:29:15Z

@lordofhyphens
Thx for fast reply, in regard of the rule of thumb. This is ok. I think, relying on the nozzle diameter as limiting factor is reasonable, but using the nozzle diameter for not user controllable calculations is worth thinking of.

In regard of the formula above to calculate the support layer distance I would suggest to change it to min(layer_height*2,nozzle_diameter * 0.75). Because you should not assume a 0.75 * nozzle_diameter is always the best solution for every printer. (Or at least make the "security factor" 0.75 user configurable).

@alexrj & @lordofhyphens
I implemented my change mentioned in the entry post in my local code and it works perfectly as I need a 0.15mm spacing for 0.1mm layers with a .4mm nozzle. In the original version I always ended with 0.4mm or more.
I think, this is just a bug not found for a long time, as a lot of people using slic3r try to use no support or rely on meshmixer as myself.

lordofhyphens · 2016-06-26T15:11:20Z

Yeah, it seemed odd. What I can do is put together a PR for that change and put it up for more widespread testing. The other change to have it take the min of the layer height + extra for non soluble support also seems like a fair plan, but it definitely is something that would need testing. I will add it to the PR.

…diameter+extra. Addresses slic3r#3376

lordofhyphens · 2016-06-26T16:07:31Z

Looking over the other change I think it needs some more experimenting or testing with different layer heights, so I'll put a separate PR in for it. The comments indicate that the max() is used to avoid inserting a ton of ultra-thin support layers; that is, it defaults to as thick as layers as it can manage for the non-contact areas.

Moreover, slic3r won't even let you choose layer heights > nozzle diameter to begin with. If you're having flow issues at 75% of the stated nozzle width I'm wondering if you're lying to slic3r about the nozzle width (in which case you deserve what you get ;) )

Ricky1975 · 2016-06-26T21:01:56Z

@lordofhyphens
Of course, the adaption of layer heights need some testing, as it only is an idea. Yes, I have flow issues if using .75 * nozzle_diameter (without lying :-) ) I create technical designs and this depends on mechanical stiffness. I also do not care of printing time very much. Quality is the key for me (and therefore I switched so slic3r as nearly everything can be configured by hand).
So this is no issue of slic3r itself but an issue of my quality expectations - and as slic3r is the best tool with the most user tweak able settings, I suggest to make this factor tweak able, too. Maybe by just making the 0.75 security factor tweak able by the user for all calculations. I would prefer .5 for my personal usage.
Thx for adding the other fix - I hope it will find it's way into the official code soon :-)

lordofhyphens · 2016-06-28T15:07:48Z

@Ricky1975 #3377.

There's a win32 package for that and other things at https://github.com/alexrj/Slic3r/files/336282/slic3r-most_prs.zip

lordofhyphens · 2016-06-29T19:09:16Z

The PR needs a bit more work I think though; I can just see a bunch of users complaining that their support contact distance is too close now with this PR.

Ricky1975 · 2016-06-30T18:50:07Z

First of all - I don't see the complaints - please tell me where.
On the other hand, of course this will happen if they do not change their values to reasonable settings (default). If they used something like 0.00001 before to minimize the distance (just keep extra above 0) and tweaked their nozzle size to .25 they got a distance of about 0.25mm before and will now get 0.00001 (extra) + 0.1mm (layerheight).
So everybody complaining should reset their values to reasonable settings (just an idea where this error might come from, because we should not assume, everybody understands the change but the setting are kept).

lordofhyphens · 2016-06-30T19:12:01Z

I have learned to not assume that people are easy-going about changes is all :) The PR is going to get merged in regardless (well barring something catastrophic coming of it) though and we will have to update documents and the wiki. I would appreciate it if people testing this did print tests too with it.

lordofhyphens · 2016-06-30T19:27:53Z

I think what I am going to do for the default value is change it to 0.4; in a lot of use cases this should yield similar results to 0.2 with the old setup. I will also probably change the ui to be a straight number in mm with a tooltip stating that it is an offset from the previous layer, and that 0 is soluable. Probably could use some people doing some SCIENCE!! to work out some sane defaults based on layer height.

Mikk36 · 2016-06-30T19:33:55Z

Did some prints today, worked just as expected. Finding a good value is an art though, but that's how it's supposed to be, there is no single best value.

…diameter+extra. Addresses slic3r#3376

lordofhyphens · 2016-07-11T06:27:38Z

@alexrj I would appreciate your input on this problem. I think that the addition of the nozzle diameter to the extra when non-soluable does constitute a UI bug at minimum and causes failures at smaller layer heights.

The referenced PR removes the special casing, permitting support separation values 0> and <= nozzle_diameter.

Martronic-SA · 2016-09-15T12:15:52Z

Ok I better understand why the support material doesn't follow my settings!
When I'm printing 0.1mm layer height with distance set to 0.1mm, the real support distance is always much greater and the upper part doesn't rely on the support.
If I set 0.1 mm, I was expecting to get 0.1mm even if it's not enough and my piece sticks to the support.

lordofhyphens · 2016-09-15T13:06:26Z

Try the branch referenced on the pull request @Martronic-SA

VanessaE · 2017-01-15T03:06:34Z

There hasn't been much movement on this PR; would this also fix the case where the actual contact distance is affected by the angle of the supported face? If so, can we please get this merged to master?

Seems like no matter what I do, I can't get a consistent contact distance between support and object, even at routine 0.1mm layer height. Even a distance of 0 doesn't result in full contact if the supported face is too far beyond the configured overhang threshold (I think).

The wings in this model, http://www.thingiverse.com/thing:824995 , when turned with the concave side (that is, the bottom side) of the wings facing down, will expose this issue pretty readily.

bubnikv · 2017-01-15T08:31:16Z

There are two types of contacts. Top and bottom. Which one are you referring to? I am currently tuning the new supports. I am getting a good top contact layer (the one below the overhangs), but the bottom contact layer is not that easy to remove. I am printing the test objects side by side, sliced by the Slic3r and simplify3d. The simplify3d has their supports tuned very well. The pla supports remove in a single piece without leaving a scar. It works like a magic. That is the benchmark I would like to reach.

…

On Jan 15, 2017 4:06 AM, "Vanessa Ezekowitz" ***@***.***> wrote: There hasn't been much movement on this PR; would this also fix the case where the actual contact distance is affected by the angle of the supported face? Seems like no matter what I do, I can't get a consistent contact distance between support and object, even at routine 0.1mm layer height. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#3376 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFj5I1AWsCscKFeISsT2S_r4d7oJCOCuks5rSY09gaJpZM4I-WQj> .

VanessaE · 2017-01-15T08:37:03Z

I was referring to top contacts.

bubnikv · 2017-01-15T09:08:32Z

Just to make sure. You mean the contacts supporting overhangs. What material? That makes a lot of difference.Pla cracks away easily, pet is worse, abs is very difficult as the contact layer welds firmly no matter what you do and how large the distance is, as long as the cooling fan is off. With abs my impression is, you better use a sparse interface. By the way, I developed couple of blisters on my hands over the last couple of days just by printing multiple models and removing the supports myself. It gives one quite an idea about the usability of the supports. I think you requested a feature for soluble supports to print a non-soluble interface layer below the first soluble layer. We are having the same problem sticking PVA over pla with our multi material system, so I will likely have to develop something.

…

On Jan 15, 2017 9:37 AM, "Vanessa Ezekowitz" ***@***.***> wrote: I was referring to top contacts. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#3376 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFj5I1mQOV5_6ZWAL6SxHYD-aFvtmH4iks5rSdqygaJpZM4I-WQj> .

VanessaE · 2017-01-15T09:13:56Z

Yes, I meant exactly that.

…diameter+extra. Addresses slic3r#3376

lordofhyphens mentioned this issue Jun 26, 2016

Feature Request: combine support material every X layers #3377

Open

lordofhyphens added Needs testing NEED ADVICE AND IDEAS FROM COMMUNITY Support material labels Jun 26, 2016

lordofhyphens added this to the 1.3.5 milestone Jun 26, 2016

lordofhyphens added a commit to lordofhyphens/Slic3r that referenced this issue Jun 26, 2016

Moved the contact distance to layer height + extra instead of nozzle …

00742c7

…diameter+extra. Addresses slic3r#3376

lordofhyphens mentioned this issue Jun 26, 2016

Changed contact distance to layer height + extra #3379

Closed

lordofhyphens added the Fixed with PR available to merge There is an update to address this issue in an open pull request. label Jun 29, 2016

lordofhyphens added a commit to lordofhyphens/Slic3r that referenced this issue Jul 9, 2016

Moved the contact distance to layer height + extra instead of nozzle …

00183ca

…diameter+extra. Addresses slic3r#3376

lordofhyphens added a commit to lordofhyphens/Slic3r that referenced this issue Apr 5, 2017

Moved the contact distance to layer height + extra instead of nozzle …

7097965

…diameter+extra. Addresses slic3r#3376

lordofhyphens mentioned this issue Apr 5, 2017

Contact Z-distance is double what is supposed to be #2930

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Support material z-distance: wrong variable used #3376

Support material z-distance: wrong variable used #3376

Ricky1975 commented Jun 25, 2016 •

edited

Loading

Ricky1975 commented Jun 25, 2016

lordofhyphens commented Jun 26, 2016 •

edited

Loading

lordofhyphens commented Jun 26, 2016

Ricky1975 commented Jun 26, 2016

lordofhyphens commented Jun 26, 2016 via email

lordofhyphens commented Jun 26, 2016

Ricky1975 commented Jun 26, 2016

lordofhyphens commented Jun 28, 2016

lordofhyphens commented Jun 29, 2016

Ricky1975 commented Jun 30, 2016

lordofhyphens commented Jun 30, 2016 via email

lordofhyphens commented Jun 30, 2016 via email

Mikk36 commented Jun 30, 2016

lordofhyphens commented Jul 11, 2016

Martronic-SA commented Sep 15, 2016

lordofhyphens commented Sep 15, 2016

VanessaE commented Jan 15, 2017 •

edited

Loading

bubnikv commented Jan 15, 2017 via email

VanessaE commented Jan 15, 2017

bubnikv commented Jan 15, 2017 via email

VanessaE commented Jan 15, 2017

Support material z-distance: wrong variable used #3376

Support material z-distance: wrong variable used #3376

Comments

Ricky1975 commented Jun 25, 2016 • edited Loading

Version

Operating system type + version

Behavior

Ricky1975 commented Jun 25, 2016

lordofhyphens commented Jun 26, 2016 • edited Loading

lordofhyphens commented Jun 26, 2016

Ricky1975 commented Jun 26, 2016

lordofhyphens commented Jun 26, 2016 via email

lordofhyphens commented Jun 26, 2016

Ricky1975 commented Jun 26, 2016

lordofhyphens commented Jun 28, 2016

lordofhyphens commented Jun 29, 2016

Ricky1975 commented Jun 30, 2016

lordofhyphens commented Jun 30, 2016 via email

lordofhyphens commented Jun 30, 2016 via email

Mikk36 commented Jun 30, 2016

lordofhyphens commented Jul 11, 2016

Martronic-SA commented Sep 15, 2016

lordofhyphens commented Sep 15, 2016

VanessaE commented Jan 15, 2017 • edited Loading

bubnikv commented Jan 15, 2017 via email

VanessaE commented Jan 15, 2017

bubnikv commented Jan 15, 2017 via email

VanessaE commented Jan 15, 2017

Ricky1975 commented Jun 25, 2016 •

edited

Loading

lordofhyphens commented Jun 26, 2016 •

edited

Loading

VanessaE commented Jan 15, 2017 •

edited

Loading