Resources covering model design guidelines
On 23 July 2018 at 14:07, Rob Ward rl.ward@bigpond.com wrote:
I am also going to try these "M3 x 3mm Solid Brass Cylinder Knurled Threaded
Round Insert Embedded Nuts" (search eBay?) as an alternative as well. I am
hoping to slip them on my soldering iron tip and insert/melt them into my
models.
That is common but I've never liked these for a few reasons:
- It's difficult to precisely place something when you're melting it
into plastic with a soldering iron. FDM printing is accurate to
~0.25mm or so, but then human hands forcing a big insert into a
too-small printed pocket with a lot of heat are less accurate. - If the hole is not easily accessible on an outside face of the part,
it's difficult to install the knurled insert. For example, on the
axles of this Kossel effector:
https://www.thingiverse.com/thing:2297083 - I'm never sure how much force I can apply to the knurled insert once
it's melted in place. Do I need to be fragile like when threading into
plastic? Will it eventually let go over time if held in tension by a
spring?
Where a captive thread force is required, modelling in a captive nut
or nyloc is far more preferable. For example, Johann's Kossel
effector: https://github.com/jcrocholl/kossel/blob/master/effector.stl
To go into a flat face, you need a small square hole to insert the
nut. You can easily find dimensions of metric ISO nuts. The hole needs
to be tight enough to stop the nut rotating. If you intend for the nut
to be removable then add a hole on the other side to poke the nut back
out with a paperclip or something.
You could also model an insertion/removal cube with more clearance and
then a hexagonal nut trap which the nut pulls back into.
If this description isn't clear, let me know and I can model it up for you.
Jamie
FWIW these two stores work well for me and also have the threaded inserts which is an alt to soldering iron push designs. Check out the rest of the h/w they sell.
- Chang-Da store:
https://www.aliexpress.com/item/100pcs-lot-Brass-insert-M2-M2-5-M3-Through-thread-brass-insert-nut-knurled-nuts-for/32840519444.html
- Zhong-Fa store:
https://www.aliexpress.com/item/100pcs-lot-M3-L-OD-4-2mm-M3-Through-hole-brass-insert-nut-knurled-nuts-for/32853029322.html
Not an affiliate link.
On 7/23/2018 4:07 PM, Rob Ward wrote:
Hi folks,
This is of interest to me as well. I have successfully threaded quite a few holes in 3-D models and found it very satisfactory. Choosing the right hole though is somewhat hit or miss for me, and I will print Hugo's plate with different holes as a better indication of what my own printer (at its settings) makes holes.
I am also going to try these "M3 x 3mm Solid Brass Cylinder Knurled Threaded Round Insert Embedded Nuts" (search eBay?) as an alternative as well. I am hoping to slip them on my soldering iron tip and insert/melt them into my models. Hugo's plate with the graded holes will help here as well.
Rob
They are good points Jamie.
My main concern was if I used the soldering iron tip to push them in
they may very well come back out on the soldering iron seeing they have
a thread that could grip the tip. I think I will look into a simple
"push" fit with a tool (ie a mounted M3 thread on a handle like a
screwdriver, or even a mini press) and try various hole sizes. That way
I can address your concern about the orientation better than pushing
something into sloppy melted plastic. After that I might try a slight
touch of the iron just to get the spikes to settle into the plastic
after insertion.
I have used holes with "hidden" nuts and they do work well.
Rob
On 23/07/18 15:21, Jamie Bainbridge wrote:
On 23 July 2018 at 14:07, Rob Ward rl.ward@bigpond.com wrote:
I am also going to try these "M3 x 3mm Solid Brass Cylinder Knurled Threaded
Round Insert Embedded Nuts" (search eBay?) as an alternative as well. I am
hoping to slip them on my soldering iron tip and insert/melt them into my
models.
That is common but I've never liked these for a few reasons:
- It's difficult to precisely place something when you're melting it
into plastic with a soldering iron. FDM printing is accurate to
~0.25mm or so, but then human hands forcing a big insert into a
too-small printed pocket with a lot of heat are less accurate. - If the hole is not easily accessible on an outside face of the part,
it's difficult to install the knurled insert. For example, on the
axles of this Kossel effector:
https://www.thingiverse.com/thing:2297083 - I'm never sure how much force I can apply to the knurled insert once
it's melted in place. Do I need to be fragile like when threading into
plastic? Will it eventually let go over time if held in tension by a
spring?
Where a captive thread force is required, modelling in a captive nut
or nyloc is far more preferable. For example, Johann's Kossel
effector: https://github.com/jcrocholl/kossel/blob/master/effector.stl
To go into a flat face, you need a small square hole to insert the
nut. You can easily find dimensions of metric ISO nuts. The hole needs
to be tight enough to stop the nut rotating. If you intend for the nut
to be removable then add a hole on the other side to poke the nut back
out with a paperclip or something.
You could also model an insertion/removal cube with more clearance and
then a hexagonal nut trap which the nut pulls back into.
If this description isn't clear, let me know and I can model it up for you.
Jamie
I make a lot of things with heatfit inserts and they work very well for me.
They don't look like the Aliexpress ones above that are intended for
injection moulding. They start with a plain section that is slightly
smaller than the hole you insert them into, so they are guided all the way
down and don't drift.
https://www.insertsdirect.com/acatalog/heatfit-insert.html
I use a soldering iron set to 200C (which is a lot less than for soldering)
with a conical bit that hasn't even been used for solder. I hold the iron
for a few seconds before giving a firm push. I twist anti-clockwise before
pulling back to stop the bit catching in the thread.
I have seen Warner Berry use a custom turned bit with a shoulder and the
iron mounted in a drill press to keep it all parallel. That is probably the
best way to do it, but as I use four different sizes I prefer the conical
bit.
Using polyholes http://hydraraptor.blogspot.com/2011/02/polyholes.htmlI
get better than 0.25mm accuracy with FDM, close to 0.1mm. Good enough for
heatfit, tapping, press fit and clearance holes. For higher precision like
pulley bores I print them undersized and then ream with an H7 reamer in a
drill press set to low speed. That gives a nice snug fit on a motor shaft.
When you drill or ream soft plastic holes come out very slightly
undersized, so a reamed plastic pulley is a little tighter than an
aluminium one. However when you get to such tight tolerances as H7
temperature has a significant effect on the size.
On 23 July 2018 at 07:26, Rob Ward rl.ward@bigpond.com wrote:
They are good points Jamie.
My main concern was if I used the soldering iron tip to push them in they
may very well come back out on the soldering iron seeing they have a thread
that could grip the tip. I think I will look into a simple "push" fit with
a tool (ie a mounted M3 thread on a handle like a screwdriver, or even a
mini press) and try various hole sizes. That way I can address your concern
about the orientation better than pushing something into sloppy melted
plastic. After that I might try a slight touch of the iron just to get the
spikes to settle into the plastic after insertion.
I have used holes with "hidden" nuts and they do work well.
Rob
On 23/07/18 15:21, Jamie Bainbridge wrote:
On 23 July 2018 at 14:07, Rob Ward rl.ward@bigpond.com wrote:
I am also going to try these "M3 x 3mm Solid Brass Cylinder Knurled
Threaded
Round Insert Embedded Nuts" (search eBay?) as an alternative as well. I
am
hoping to slip them on my soldering iron tip and insert/melt them into my
models.
That is common but I've never liked these for a few reasons:
- It's difficult to precisely place something when you're melting it
into plastic with a soldering iron. FDM printing is accurate to
~0.25mm or so, but then human hands forcing a big insert into a
too-small printed pocket with a lot of heat are less accurate. - If the hole is not easily accessible on an outside face of the part,
it's difficult to install the knurled insert. For example, on the
axles of this Kossel effector:
https://www.thingiverse.com/thing:2297083 - I'm never sure how much force I can apply to the knurled insert once
it's melted in place. Do I need to be fragile like when threading into
plastic? Will it eventually let go over time if held in tension by a
spring?
Where a captive thread force is required, modelling in a captive nut
or nyloc is far more preferable. For example, Johann's Kossel
effector: https://github.com/jcrocholl/kossel/blob/master/effector.stl
To go into a flat face, you need a small square hole to insert the
nut. You can easily find dimensions of metric ISO nuts. The hole needs
to be tight enough to stop the nut rotating. If you intend for the nut
to be removable then add a hole on the other side to poke the nut back
out with a paperclip or something.
You could also model an insertion/removal cube with more clearance and
then a hexagonal nut trap which the nut pulls back into.
If this description isn't clear, let me know and I can model it up for
you.
Jamie
On 07/23/2018 01:21 AM, Jamie Bainbridge wrote:
melting it into plastic with a soldering iron
I set the hole OD to the insert knurl OD, (optionally) butter the knurl
with epoxy, use a screw as pilot, and ram it into the hole with a drill
press to keep it perpendicular.
Dry, showing the drill press:
https://softsolder.com/2016/03/07/vacuum-tube-leds-noval-tube-on-a-platter/
An early attempt with epoxy:
https://softsolder.com/2016/03/14/knurled-inserts-epoxy-anchoring/
The 3D printed hole will be at the right location, so using epoxy
eliminates melting-plastic distortion / misalignment and ensures great
bonding to the rough hole walls.
I tried adding recessed rings to the hole for more bonding surface, but,
given the overall lack of strength, that seemed like gilding gold.
Tedious if you're in production, no big deal for my onesie-twosie parts.
--
Ed
https://softsolder.com
Each to his own, but the ones I use are designed to be hot pressed into
thermoplastic and the correct hole size is specified. I have never had one
fail or more out of alignment enough to matter. Nearly everything I print
nowadays uses them.
On 23 July 2018 at 14:54, Ed Nisley ed.nisley@pobox.com wrote:
On 07/23/2018 01:21 AM, Jamie Bainbridge wrote:
melting it into plastic with a soldering iron
I set the hole OD to the insert knurl OD, (optionally) butter the knurl
with epoxy, use a screw as pilot, and ram it into the hole with a drill
press to keep it perpendicular.
Dry, showing the drill press:
https://softsolder.com/2016/03/07/vacuum-tube-leds-noval-tub
e-on-a-platter/
An early attempt with epoxy:
https://softsolder.com/2016/03/14/knurled-inserts-epoxy-anchoring/
The 3D printed hole will be at the right location, so using epoxy
eliminates melting-plastic distortion / misalignment and ensures great
bonding to the rough hole walls.
I tried adding recessed rings to the hole for more bonding surface, but,
given the overall lack of strength, that seemed like gilding gold.
Tedious if you're in production, no big deal for my onesie-twosie parts.
--
Ed
https://softsolder.com
OpenSCAD mailing list
Discuss@lists.openscad.org
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org
This is great info. You should write a blog post about this so it's
not lost in the annals of historic forum posts which few will ever
read.
On 23 July 2018 at 18:43, nop head nop.head@gmail.com wrote:
I make a lot of things with heatfit inserts and they work very well for me.
They don't look like the Aliexpress ones above that are intended for
injection moulding. They start with a plain section that is slightly smaller
than the hole you insert them into, so they are guided all the way down and
don't drift. https://www.insertsdirect.com/acatalog/heatfit-insert.html
I use a soldering iron set to 200C (which is a lot less than for soldering)
with a conical bit that hasn't even been used for solder. I hold the iron
for a few seconds before giving a firm push. I twist anti-clockwise before
pulling back to stop the bit catching in the thread.
I have seen Warner Berry use a custom turned bit with a shoulder and the
iron mounted in a drill press to keep it all parallel. That is probably the
best way to do it, but as I use four different sizes I prefer the conical
bit.
Using polyholes I get better than 0.25mm accuracy with FDM, close to 0.1mm.
Good enough for heatfit, tapping, press fit and clearance holes. For higher
precision like pulley bores I print them undersized and then ream with an H7
reamer in a drill press set to low speed. That gives a nice snug fit on a
motor shaft. When you drill or ream soft plastic holes come out very
slightly undersized, so a reamed plastic pulley is a little tighter than an
aluminium one. However when you get to such tight tolerances as H7
temperature has a significant effect on the size.
On 23 July 2018 at 07:26, Rob Ward rl.ward@bigpond.com wrote:
They are good points Jamie.
My main concern was if I used the soldering iron tip to push them in they
may very well come back out on the soldering iron seeing they have a thread
that could grip the tip. I think I will look into a simple "push" fit with a
tool (ie a mounted M3 thread on a handle like a screwdriver, or even a mini
press) and try various hole sizes. That way I can address your concern about
the orientation better than pushing something into sloppy melted plastic.
After that I might try a slight touch of the iron just to get the spikes to
settle into the plastic after insertion.
I have used holes with "hidden" nuts and they do work well.
Rob
On 23/07/18 15:21, Jamie Bainbridge wrote:
On 23 July 2018 at 14:07, Rob Ward rl.ward@bigpond.com wrote:
I am also going to try these "M3 x 3mm Solid Brass Cylinder Knurled
Threaded
Round Insert Embedded Nuts" (search eBay?) as an alternative as well. I
am
hoping to slip them on my soldering iron tip and insert/melt them into
my
models.
That is common but I've never liked these for a few reasons:
- It's difficult to precisely place something when you're melting it
into plastic with a soldering iron. FDM printing is accurate to
~0.25mm or so, but then human hands forcing a big insert into a
too-small printed pocket with a lot of heat are less accurate. - If the hole is not easily accessible on an outside face of the part,
it's difficult to install the knurled insert. For example, on the
axles of this Kossel effector:
https://www.thingiverse.com/thing:2297083 - I'm never sure how much force I can apply to the knurled insert once
it's melted in place. Do I need to be fragile like when threading into
plastic? Will it eventually let go over time if held in tension by a
spring?
Where a captive thread force is required, modelling in a captive nut
or nyloc is far more preferable. For example, Johann's Kossel
effector: https://github.com/jcrocholl/kossel/blob/master/effector.stl
To go into a flat face, you need a small square hole to insert the
nut. You can easily find dimensions of metric ISO nuts. The hole needs
to be tight enough to stop the nut rotating. If you intend for the nut
to be removable then add a hole on the other side to poke the nut back
out with a paperclip or something.
You could also model an insertion/removal cube with more clearance and
then a hexagonal nut trap which the nut pulls back into.
If this description isn't clear, let me know and I can model it up for
you.
Jamie
I'd probably just make it easy for myself and simply put a dimple where the
hole should be, then drill it afterwards. That way, you'll get precise
dimensions.
--
Sent from: http://forum.openscad.org/
On 27 July 2018 at 01:44, Troberg troberg.anders@gmail.com wrote:
I'd probably just make it easy for myself and simply put a dimple where the
hole should be, then drill it afterwards. That way, you'll get precise
dimensions.
Then you don't get perimeters around the hole, so the part would be
weak against the compression of the fastener.
To avoid this you could use the part modifier function of the slicer
to print 100% infill just around that drill area. Both Cura and Slic3r
have the ability to do this. S3D does not afaik, I think it can only
do multiple processes based on layers not areas, maybe I am wrong.
You also would need a drill press to get straight holes through thick
parts. That would matter if you're mating to something else precisely
made, such as using an M3x50mm screw to hold a NEMA17 motor against a
thick part.
Jamie
Just an update.
Based on some of the information I got here I created a model for a part
attempting to duplicate an existing part printed from a 3rd party model.
Long story short using ShapeWays and Versatile Plastic (which was used on
the other part) the result was spot on and the holes and hexnut recesses
were most satisfactory, no work needed, The part could be used just as it
was printed.
Except to use the part I need two more.
Frederick
--
Sent from: http://forum.openscad.org/