This is very helpful, both of you. I am looking at metric machine screws only and self tapping in plastic only. Subtracting the pitch seems like a nice easy heuristic. My background is strictly software, not mech engineering so this is all new to me. So far I have only worked with PLA.

I’m going to put together a little test print to check exactly what tolerances are needed to cleanly drop nuts into holes, etc. I just finished doing the BOSL2 test print for $slop. Coincidently I randomly picked t = .1 for my tolerance amount when I first started. I usually use 2*t for a tab that must insert into a lot and that has worked for me well enough. Oddly the $slop for my printer turns out to be … .1 mm.

-Bob
Tucson AZ

On Nov 3, 2021, at 04:05, nop head <nop.head@gmail.com mailto:nop.head@gmail.com> wrote:

For metric screws the minor diameter is not nominal minus sqrt(3) times the pitch. That is the base of the triangle but it is fillited at 5/8 and 1/8. According to Wikipedia it is the major diameter minus 1.082532 times the pitch. Using standard drill sizes, that is close enough to 1 times the pitch.

<image.png>

<image.png>

On Wed, 3 Nov 2021 at 10:53, Adrian Mariano <avm4@cornell.edu mailto:avm4@cornell.edu> wrote:
He didn't say he wanted to do tapped machine screws. So using OD minus
pitch means you're making an assumption about the desired amount of
extra for tapping.  The nominal ID will be the diameter mus
sqrt(3)*pitch so by subtracting only the pitch you are implicitly
making the hole larger than the screw ID by sqrt(3)-1 times the pitch,
or about .7 times the pitch.  Is that always the right amount of
"extra" to leave for tapping?  I don't know the answer.  This rule of
thumb will work the same for UTS screws because they have the same
geometry.

On Wed, Nov 3, 2021 at 4:41 AM nop head <nop.head@gmail.com mailto:nop.head@gmail.com> wrote:

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org mailto:discuss-leave@lists.openscad.org

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org mailto:discuss-leave@lists.openscad.org

-Bob
Tucson AZ

Tolerance parameters are ignored because one doesn't know what they are and
it would be a pain to deal with them.  But since the original poster said
he was using BOSL2 and it provides screw dimensions with the actual
tolerances considered, why not use the real screw dimensions?

We were talking about holes, meaning nuts, not screws.  An M6 nut must have
an OD larger than its nominal size or it's not ISO compliant.  I'm not sure
how you measure the OD of a nut.  For machine screws the 6g tolerance
specifies a diameter between 5.797 mm and 5.974mm, but the 8g tolerance
would allow your 5.75mm screws.  The dimensions of screws and nuts have to
be different or they wouldn't be able to screw together.

Are wood screws tightly specified, meaning that if I gather #8 wood screws
from 100 manufacturers they are all the same, with the same thread form and
dimensions?  I don't see how that can be the case if there isn't a standard
document that we can read that tells us what the wood screw dimensions
are.  I see wood screw manufacturers bragging about their superior threads
or other features, which make it sound like everybody is out there doing
their own thing.

Note that for small wood screws like #4 and below, I believe, it is
stronger to tap and use a machine screw than it is to use a wood screw.
I'm not sure exactly at what point that changes, and I imagine it might
depend on the wood species.  When I installed my woodworking vise all the
hardware was attached to the wood with tapped holes in the wood using
machine screws (which was how the vise manufacturer specified that it be
done) even for #8 screws.  Of course, another advantage of tapping wood for
machine screws is that you can remove and reinstall the item without loss
of holding power, which may not be true for regular wood screws.

On Thu, Nov 4, 2021 at 7:27 AM nop head nop.head@gmail.com wrote:

On Thursday 04 November 2021 14:20:42 Bob Carlson wrote:

I'm gonna get a bit opinionated here. I'm too old at 87 to waste time
argueing.

rant mode on:

PLA is too brittle to self tap into, generally speaking, so if using PLA
I'd consider the hole to be a pilot hole for the proper sized drill, and
then use a quality real carbide cutting tap and faith. You will need
lots of that.

PETG is an entirely different horse. So use PETG if putting threads into
a hole by any means.

Most printers with stock printheads that take the bowden tube clear thru
to the rear of the nozzle will not live long enough to find the optimum
settings for PETG, the capricorn slowly goes away at the 235C+ temps
PETG needs and leakage will destroy your prints by ripping them off the
build plate when the head comes by to lay the next layer and hits the
blob that dripped off the head onto the print and cooled there.

The BIQU BX's H2 head supposedly has a kit that will goto 500C, but no
one is stocking it or listing it for order including BIQU. I've 3 of
those H2 heads I'd like to buy that kit for, all failed and leaking like
a collander by the third print. The only other printer that eats PETG
over the long haul, and I've a room full of them from several makers
that claimed PETG compatability, is the Prusa MK3S kit.

So save your sheckels, get one and spend 3 days building the kit if you
want to use PETG on an everyday basis.

Look for printers that claim to have an E3D or better hot end. They don't
come stock on $150, or even $500 printers, not even on the nearly $2000
Dremel 3d45. The $800 Prusa kit does. With a titanium heat break, it
Just Works as long as you clean the hot nozzle with a brass toothbrush
to start the job every time.

/rant mode off.

mailto:nop.head@gmail.com> wrote:

mailto:bob@rjcarlson.com> wrote:

Cheers, Gene Heskett.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis
  Genes Web page http://geneslinuxbox.net:6309/gene

On Thursday 04 November 2021 16:50:23 Adrian Mariano wrote:

Absolutely correct. if 100 makers, all will have their own private specs,
if only to get around O.P. patents which there must be hundreds of.

It does, with clear maple and ash being good, mahogany nearly so while
oak might split on a grain boundary. The lower density stuff is often
better off just glued since the glue can penetrate the pores of bass,
balsa or butternut quite easily. But TBT, the only thing I've used
butternut for is as a slightly lighter colored walnut in kitchen cabinet
door panels. No screws there as they slide into slots in the door frame
stiles in otherwise BBLB cabinets in a 1970's national catalog house.
Butternut is actually a kissing cousin of american black walnut, but
about half the weight per cubic foot.

Thats how I did mine too, in std pine legs & 2x4 top, etc.

Unless, when replaceing the same screw in the same hole, the screw is
rotated backwards until it can be felt to have dropped into the old
threads.  Sometimes you can get away with it if the new screw is from
the same box as the old one.

And requires flexablity in the material surrounding the hole, something
sorely missing from the PLA for printers.

[...]

Cheers, Gene Heskett.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis
  Genes Web page http://geneslinuxbox.net:6309/gene

I have only used self tap screws a couple of times since I noticed how someone else had used them. I use them for parts I do not ever expect to have to disassemble. Perhaps I got lucky with my choice of diameter, but they’ve worked fine in PLA for me so far. They have been M3 or M4. I also put in a 1 mm “pilot” that is full nominal OD.

-Bob
Tucson AZ

On Nov 4, 2021, at 15:26, Gene Heskett gheskett@shentel.net wrote:

On Thursday 04 November 2021 16:50:23 Adrian Mariano wrote:

Absolutely correct. if 100 makers, all will have their own private specs,
if only to get around O.P. patents which there must be hundreds of.

It does, with clear maple and ash being good, mahogany nearly so while
oak might split on a grain boundary. The lower density stuff is often
better off just glued since the glue can penetrate the pores of bass,
balsa or butternut quite easily. But TBT, the only thing I've used
butternut for is as a slightly lighter colored walnut in kitchen cabinet
door panels. No screws there as they slide into slots in the door frame
stiles in otherwise BBLB cabinets in a 1970's national catalog house.
Butternut is actually a kissing cousin of american black walnut, but
about half the weight per cubic foot.

Thats how I did mine too, in std pine legs & 2x4 top, etc.

Unless, when replaceing the same screw in the same hole, the screw is
rotated backwards until it can be felt to have dropped into the old
threads.  Sometimes you can get away with it if the new screw is from
the same box as the old one.

And requires flexablity in the material surrounding the hole, something
sorely missing from the PLA for printers.

[...]

Cheers, Gene Heskett.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis
  Genes Web page http://geneslinuxbox.net:6309/gene

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org

On Thursday 04 November 2021 19:43:32 Bob Carlson wrote:

That in PLA, might be the salvation. Bears trying for sure. If thick
enough, 2mm deep, on both ends of the hole, anything to relieve the
surface stress where the cracks always start.

wrote:

Cheers, Gene Heskett.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis
  Genes Web page http://geneslinuxbox.net:6309/gene

OK, I have done some test prints of holes for M5 nuts. Nominally these are supposed to be 8mm*4mm. The measured diameter of my nuts is around 7.85.

When the hole is vertical, ie the dimensions of the hole are in the XY plane, nominal plus .05 seems to work with a snug fit and nominal plus .1 definitely works.

If the 8mm dimension is parallel to the Z axis, then nominal plus .25 works, with nominal + .2 giving a very tight fit. To be clear the “top” of the hole is a roof.

I can see possibilities here. One is that the “roof” sags and causes the difference. This might be unavoidable if true. It’s only 4 mm long though.
A second is that my printer is not calibrated well enough so Z dimensions are off a bit.
Third is that the Z dimension is inherently less accurate in this kind of printer (MK3S+) so I just have to deal with it. Unavoidable again.

Thoughts?

Thanks!

-Bob
Tucson AZ

On Nov 4, 2021, at 11:20, Bob Carlson bob@rjcarlson.com wrote:

This is very helpful, both of you. I am looking at metric machine screws only and self tapping in plastic only. Subtracting the pitch seems like a nice easy heuristic. My background is strictly software, not mech engineering so this is all new to me. So far I have only worked with PLA.

I’m going to put together a little test print to check exactly what tolerances are needed to cleanly drop nuts into holes, etc. I just finished doing the BOSL2 test print for $slop. Coincidently I randomly picked t = .1 for my tolerance amount when I first started. I usually use 2*t for a tab that must insert into a lot and that has worked for me well enough. Oddly the $slop for my printer turns out to be … .1 mm.

-Bob
Tucson AZ

On Nov 3, 2021, at 04:05, nop head <nop.head@gmail.com mailto:nop.head@gmail.com> wrote:

For metric screws the minor diameter is not nominal minus sqrt(3) times the pitch. That is the base of the triangle but it is fillited at 5/8 and 1/8. According to Wikipedia it is the major diameter minus 1.082532 times the pitch. Using standard drill sizes, that is close enough to 1 times the pitch.

<image.png>

<image.png>

On Wed, 3 Nov 2021 at 10:53, Adrian Mariano <avm4@cornell.edu mailto:avm4@cornell.edu> wrote:
He didn't say he wanted to do tapped machine screws. So using OD minus
pitch means you're making an assumption about the desired amount of
extra for tapping.  The nominal ID will be the diameter mus
sqrt(3)*pitch so by subtracting only the pitch you are implicitly
making the hole larger than the screw ID by sqrt(3)-1 times the pitch,
or about .7 times the pitch.  Is that always the right amount of
"extra" to leave for tapping?  I don't know the answer.  This rule of
thumb will work the same for UTS screws because they have the same
geometry.

On Wed, Nov 3, 2021 at 4:41 AM nop head <nop.head@gmail.com mailto:nop.head@gmail.com> wrote:

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org mailto:discuss-leave@lists.openscad.org

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org mailto:discuss-leave@lists.openscad.org

-Bob
Tucson AZ

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org

For most nuts I use cylinder(h=t+tl,d=(a*1.154)+ td,$fn=6); where  td =
0.3 and tl is 0.8 and for the 5mm nuts I have then af is 7 and t is 4.
works out to be good enough to insert nuts in any orientation in middle
of printing  Hole sizes will be different between horizontal and
vertical holes of same diameter, but iirc Cura has adjustments for that,
not that I bother. Depending on the filament direction, the 'roof' may
be different, but for 5mm nut it should not matter either way. Provided
the nut doesn't turn in the hole, it's good enough since  I would never
use my 3d printing for serious structural parts where torque settings
for nuts may be needed.

On 06/11/2021 22:45, Bob Carlson wrote:

My library has the nut radius for M5 as 9.2mm, which gives 7.97mm across
the flats. I measure 7.87mm.

I have a nut trap module that makes the hexagonal hole exactly the same as
the nut when it is vertical and adds layer height / 4 when horizontal. That
has always worked for me.

When horizontal the layer above is always a bit lower unless printed with
support because the filament is round when stretched across a bridge but
flat when not.

On Sat, 6 Nov 2021 at 23:48, Ray West raywest@raywest.com wrote:

oops, just checked - I use 7 for 4mm, 8 for 5mm. ( my comment at the
head of my library was aaf, but the values I used were not.) corrected
now, thanks.

// diam circle = af/0.866
    // 3mm af 5.5, t 2.5
    // 4mm af 7, t 3
    // 5mm af 8, t 4
    // 6mm af 10, t 5
    // 8mm af 13, t 6.5
    // 10mm af 17 t 7.5
    // 12mm af 19 t 9.5
 //allow tolerance of 0.3 on diameters for nut and 0.8 on length
// tolerance of 0.5 on diameter of bolt

On 07/11/2021 01:04, nop head wrote: