Axle hole primitive on a Bar connection.

[Magnus Forsberg]

This is a valid connection, used in official BI

1 0 0 -8 0 1 0 0 0 1 0 0 0 1 20482.dat
1 72 0 -18 0 1 0 0 0 1 0 0 0 1 2723.dat

How many of the axle hole primitives are allowed to attach on a Bar sized cylinder?
Digging in my parts I found only a few, mainly older parts that didn't fit on a Bar. Is the design of the axlehol primitive correct?

I was asked to maybe change the design on the new wheelholder brick, 65634 , but I can't find a better suited axle primitive.
The best suited primitive is maybe axl3hol6, but it doesn't look right.

Are we missing an axle primitive?

[Philippe Hurbain]

How many of the axle hole primitives are allowed to attach on a Bar sized cylinder?
Digging in my parts I found only a few, mainly older parts that didn't fit on a Bar. Is the design of the axlehol primitive correct?

If you mean that bars collide with axlehole primitives, yes they do (and always did... don't prevent me sleeping!).
Otherwise, AFAIK, all axleholes - even very old - fit bars (just tested again with toothed toggle connectors (4273)

[Orion Pobursky]

If you mean that bars collide with axlehole primitives, yes they do (and always did... don't prevent me sleeping!).
Otherwise, AFAIK, all axleholes - even very old - fit bars (just tested again with toothed toggle connectors (4273)

I also believe that LDCad allows this as a legal part snapping connection. The overlap isn't a big deal and will always be a limitation of digital vs. physical bricks.

[Roland Melkert]

I also believe that LDCad allows this as a legal part snapping connection. The overlap isn't a big deal and will always be a limitation of digital vs. physical bricks.

I included 4 ldu pen into 6ldu axle hole because antenna's fit into axle holes in real life too.

In ldraw the overlap seems very big though.

[Magnus Forsberg]

In ldraw the overlap seems very big though.

Maybe it is a completely bonkers idea to change all the standard axlehole primitives....

   

[Orion Pobursky]

Maybe it is a completely bonkers idea to change all the standard axlehole primitives....

You just gave me a headache even bringing up the subject. Then we'd have to comb the library for every part that has an axle hole but doesn't use a primitive for whatever reason.

[Philippe Hurbain]

You just gave me a headache even bringing up the subject. Then we'd have to comb the library for every part that has an axle hole but doesn't use a primitive for whatever reason.

If we ever do that (and I am NOT advocating for this, scares me out of my wits!), axle should be modified too to make the sides fit 22.5° vertex of R6 circle. Not only this could make primitive substitution work better, but also would fit better the physical shape. Axle fins are 1.8mm thick (currently modeled at 1.6). Using 22.5° vertex would make it 1.836mm. Much closer...

[Magnus Forsberg]

If we ever do that (and I am NOT advocating for this, scares me out of my wits!), axle should be modified too to make the sides fit 22.5° vertex of R6 circle. Not only this could make primitive substitution work better, but also would fit better the physical shape. Axle fins are 1.8mm thick (currently modeled at 1.6). Using 22.5° vertex would make it 1.836mm. Much closer...

Ohh. That is really scary.

But if we only look at the regular axlehole.dat, my opinion is that it is doable.
There is a bunch of parts using the open primitives axlehol4 and axlehol5 that need more attention,
but most of them also use npeghol3, npeghol4 or npeghol5, so fixing these six primitives would fix a lot of parts.

But would this change, to some primitives, force all affected parts/files through the PT ?

[Philippe Hurbain]

But if we only look at the regular axlehole.dat, my opinion is that it is doable.

Possibly yes, but what kind change would you make? Blunt the sharp angle?

But would this change, to some primitives, force all affected parts/files through the PT ?

No technical need, but ideally all affected parts should be inspected for unwanted side effects...

[Orion Pobursky]

No technical need, but ideally all affected parts should be inspected for unwanted side effects...

Thats currently 60 unofficial and 233 official parts:
https://www.ldraw.org/cgi-bin/ptscan.cgi?q=axlehole

[Philippe Hurbain]

Out of curiosity, I checked in LDD... their axle hole collide with bars almost as much as LDraw ones

[Ulrich Röder]

"I was asked to maybe change the design on the new wheelholder brick, 65634 , but I can't find a better suited axle primitive.
The best suited primitive is maybe axl3hol6, but it doesn't look right. "

... what speaks against ..unorthodox tinkering a subpart in this case? ... would be at least a solution.

[Javier Orquera]

Hello everyone:

I made the primitive axl5hole.dat and have issues with the shape. I suggested a shape that I reproduce:

4 16 -2.2962 0 5.5434 -2.2962 1 5.5434 0 1 6 0 0 6
4 16 -2.45 0 3.6 -2.2962 0 4.5 -2.2962 0 5.5434 -4.2426 0 4.2426
4 16 -2.45 0 3.6 -4.2426 0 4.2426 -3.6 0 2.45 -2.8284 0 2.8284
4 16 -5.5434 0 2.2962 -4.5 0 2.2962 -3.6 0 2.45 -4.2426 0 4.2426
4 16 -2.2962 1 4.5 -2.2962 1 5.5434 -2.2962 0 5.5434 -2.2962 0 4.5
4 16 -2.45 1 3.6 -2.2962 1 4.5 -2.2962 0 4.5 -2.45 0 3.6
4 16 -2.45 1 3.6 -2.45 0 3.6 -2.8284 0 2.8284 -2.8284 1 2.8284
4 16 -2.8284 1 2.8284 -2.8284 0 2.8284 -3.6 0 2.45 -3.6 1 2.45
4 16 -3.6 0 2.45 -4.5 0 2.2962 -4.5 1 2.2962 -3.6 1 2.45
4 16 -4.5 1 2.2962 -4.5 0 2.2962 -5.5434 0 2.2962 -5.5434 1 2.2962
4 16 -2.2962 1 5.5434 -2.2962 1 4.5 -2.45 1 3.6 -4.2426 1 4.2426
4 16 -3.6 1 2.45 -4.2426 1 4.2426 -2.45 1 3.6 -2.8284 1 2.8284
4 16 -3.6 1 2.45 -4.5 1 2.2962 -5.5434 1 2.2962 -4.2426 1 4.2426
4 16 -6 0 0 -6 1 0 -5.5434 1 2.2962 -5.5434 0 2.2962


It's just a quarter circle. My intension is to do it as close as posible to the real shape.

Magfors give their suggestion, and there is the https://www.ldraw.org/cgi-bin/ptdetail.c...lehol8.dat, too.

These is why I open the thraed, to discuss the shape of my part and the complete axle primitives family.

[Magnus Forsberg]

Thank you for bringing this back to the table, but why a new thread when there's an old one about same issue?
https://forums.ldraw.org/thread-23835-po...l#pid35356

Are you suggesting that we should change not only the axlehole.dat family of primitives but all of the axle and axlehole primitives?

[Javier Orquera]

The axle hole.dat family is the priority, I think. But if the axle.dat family needs to be corrected, that too.
And if this thread is wrong, can you move it to the old one and close this? I appreciate it, thank you.

[Magnus Forsberg]

The axle hole.dat family is the priority, I think. But if the axle.dat family needs to be corrected, that too.
And if this thread is wrong, can you move it to the old one and close this? I appreciate it, thank you.

Merged the two threads into one.

[Magnus Forsberg]

Here an image trying to illustrate the old and new, Javiers and my suggestions.

   

Tan = the axle.dat primitive
Trans/white = the new primitive axl5hole.dat
Salmon = the new suggested shape from Javier
Green = My suggestion/ Willy's edit

I suggest a change only to the axle hole primitives, and that we should keep the current axle.dat as it is today.

[Willy Tschager]

• I'm fine with whatever you suggest as long as it doesn't mean that we have to recycle all parts with axle holes.
  
• I'm fine if they get an update if they are brought back to the PT for fixing other issues.
  
• I will hold every part that is re-submitted just to fix the axles.
  
• I'm fine with obsoleting the current prims and start with new ones.
  

w.

[Javier Orquera]

What do yo think on the attached file? It's based on the following https://rebrickable.com/parts/32064a/tec...ottom-pin/

Name and part number are the same as the original, since I don't know how to name and number it. Also, I did changes to the part 32064a.dat based on the image in the link.

[Magnus Forsberg]

I don't understand your suggetion. Or the article at ReBrickable.

[Javier Orquera]

I don't understand your suggetion. Or the article at ReBrickable.

The first type in Rebrickable, 32064c, in Ldraw part 32064a, have an axle hole different from the parts tracker; so I did the attached file to discuss it.

[Magnus Forsberg]

Sorry, but I still don't get it. Why would we need a axlehole primitive that is too small?
The axle doesn't fit inside your new design.

[Magnus Forsberg]

Here is an attempt to summarize what I want to do.

   

The problem is in the 3 primitives axlehole, axlehol4 and axlehol5. The inner corners should be rounded. Not 90 degrees.
But we need to keep the primitives axlehol8 - axleho10 and axleend, since those are used to make various parts with an axle.

I suggest that the 3 problem axle hole primitves are redesigned using the shape present in the axl2 and axl3 families (in a small green ring in the image). I have created a set of primitives called here axl5hol8 - axl5ho10 and axl5end (yellow, in top right corner). When the hole is given a more correct shape I need the suggested new primitives to close the affected parts.

I have made an inventory of all parts using the bad primitives and found that most of them are unaffected if the changes are made in the bad primitives, axlehole, axlehol4 and axlehol5. So, instead of making a complete set of new axl5 primitives I want to redesign the old ones.

   
(excerpts from the table)

I've found only 25 parts (14 + 6 + 5) that need further editing, mainly corrections using the suggested new(yellow) primitives.
Green in the table are possible to fix without problem. Red need further editing.
The main problem are all parts using the axleend primitives as a bottom to close the axle hole. 43 parts need editing.

All 3 problem primitives are already recycled and currently present at the Part Tracker, without a demand that all affected parts have to be recycled.

I want to edit the 3 problem primitives, allready present at the PT
I want to upload a set of new axl5 primitives.
I still need to further investigate all parts using the axlehol6 primitive.
I have started to edit all the affected parts, some are unoffical and some need to be sent to Admin.

Any objections?

[Willy Tschager]

I suggest that the 3 problem axle hole primitves are redesigned using the shape present in the axl2 and axl3 families (in a small green ring in the image). I have created a set of primitives called here axl5hol8 - axl5ho10 and axl5end (yellow, in top right corner). When the hole is given a more correct shape I need the suggested new primitives to close the affected parts.

Could we please abandon that 8.3 scheme and come up with a proper naming. It's almost 2022.

w.

[Philippe Hurbain]

Could we please abandon that 8.3 scheme and come up with a proper naming. It's almost 2022.

w.

I basically agree, but we also need to consider naming consistency with similar function primitives...

[Javier Orquera]

The problem is in the 3 primitives axlehole, axlehol4 and axlehol5. The inner corners should be rounded. Not 90 degrees.

Hello: if you want to keep the shape of the primitives, I propose to use the file axl5hole (in PT) and add an extension\adjacent\complementary "sub"part that I made:

4 16 4 1 2 4 0 2 2 0 2 2 1 2
4 16 2 0 2 2 0 4 2 1 4 2 1 2
3 16 2.8284 0 2.8284 2 0 4 2 0 2
3 16 2.8284 0 2.8284 2 0 2 4 0 2
3 16 2 1 2 2 1 4 2.8284 1 2.8284
3 16 2 1 2 2.8284 1 2.8284 4 1 2
2 24 4 1 2 2 1 2
2 24 4 0 2 2 0 2
2 24 2 1 2 2 0 2
2 24 2 0 2 2 0 4
2 24 2 1 2 2 1 4
4 16 2 0 -2 2 1 -2 2 1 -4 2 0 -4
4 16 4 1 -2 2 1 -2 2 0 -2 4 0 -2
3 16 4 0 -2 2 0 -2 2.8284 0 -2.8284
3 16 2.8284 0 -2.8284 2 0 -2 2 0 -4
3 16 2 1 -2 4 1 -2 2.8284 1 -2.8284
3 16 2 1 -2 2.8284 1 -2.8284 2 1 -4
2 24 2 1 -4 2 1 -2
2 24 2 1 -2 2 0 -2
2 24 2 0 -2 2 0 -4
2 24 2 0 -2 4 0 -2
2 24 2 1 -2 4 1 -2
4 16 -2 0 -2 -2 1 -2 -4 1 -2 -4 0 -2
4 16 -2 1 -2 -2 0 -2 -2 0 -4 -2 1 -4
3 16 -2 0 -2 -4 0 -2 -2.8284 0 -2.8284
3 16 -2 0 -2 -2.8284 0 -2.8284 -2 0 -4
3 16 -2 1 -2 -2 1 -4 -2.8284 1 -2.8284
3 16 -2 1 -2 -2.8284 1 -2.8284 -4 1 -2
2 24 -2 1 -4 -2 1 -2
2 24 -2 1 -2 -4 1 -2
2 24 -2 1 -2 -2 0 -2
2 24 -2 0 -2 -2 0 -4
2 24 -2 0 -2 -4 0 -2
4 16 -2 1 4 -2 0 4 -2 0 2 -2 1 2
4 16 -2 1 2 -2 0 2 -4 0 2 -4 1 2
3 16 -4 0 2 -2 0 2 -2.8284 0 2.8284
3 16 -2.8284 0 2.8284 -2 0 2 -2 0 4
3 16 -2 1 2 -4 1 2 -2.8284 1 2.8284
3 16 -2 1 2 -2.8284 1 2.8284 -2 1 4
2 24 -4 1 2 -2 1 2
2 24 -2 1 2 -2 1 4
2 24 -2 1 2 -2 0 2
2 24 -2 0 2 -4 0 2
2 24 -2 0 2 -2 0 4

Hope I helped in these problem.

[Magnus Forsberg]

I have started the rework of all the affected parts.

Redesign the primitives, and create a set of new primitives. Done.
2 primitives in the mail to Admin.

Rework all the affected unofficial subparts and parts. Done.
None of the new parts uploaded recently seem to be broken. Let me know if you find a broken part.

Rework all the affected official subparts and parts. Initiated  Half done Done.

[Magnus Forsberg]

454 files have been investigated.
359 was deemed unnecessary to edit. Editing the primitives will give them a correct look.

95 files was edited, due to design choices made, not using axlehole primitives.
23 files had other issues that has also been corrected/changed.
12 files found with issues unrelated to the axlehole, not fixed.

It probably still exists files affected by the changes in the axlehole primitives, that need editing. Help me find them.

[Gerald Lasser]

Do we have a new overview on the axlehole prims available now and how to use them properly?

Thanks

[Magnus Forsberg]

Do we have a new overview on the axlehole prims available now and how to use them properly?

Well, the primitive reference page needs a major overhaul and is long overdue.

I tried to "sort" the primitives into 2 groups. One group used for axles and the other for axleholes. Inside/Outside. Positive/Negative.
I could make another image, better than the one above.

And maybe we should have renamed or renumbered them? axle-primitives and axlehole-primitives?

[Gerald Lasser]

Yes, basically your picture is fine, however there are not all existing right now? correct?
what I found on PT looks like this now to me (Axle2 not with the updated shape)

   

[Magnus Forsberg]

Nae...

   

The entire axlehole family is still in use.

Yellow prims; axlehole, axlehol4, axlehol5, axlehol6 are redesigned, still using the same name.
(This made it possible to correct the design without recycling all affected parts.)

Green prims; axlehol2, axlehol3, axlehol7, axlehol8, axlehol9, axleho10, axleend are not changed.
Should only be used as axle primitives. Not as axleholes. Never inverted.

New prims; axl5hol8 , axl5hol9 , axl5ho10 , axl5end, should be used when axlehole.dat isn't possible to use.

axlehole.dat has included edgelines around the hole. This prevents it from being used inside cylindrical, recessed holes.

axl2hole.dat and axl3hole.dat doesn't. You need to add the edge primitives.

axl4hole.dat has been "autocorrected" since it is using the axlehol6 primitive. (Your image show the old design.)

And yes, updated versions of the entire axl2hole family is now present at the Part Tracker.

[Gerald Lasser]

Thanks Magnus makes it clearer!

What I would like to have is a pim of an axlehole pocket.

because currently this is a construction of:

- end (optional)
- hol2 (bottom)
- hol9 (bottom)
- hol8
- hol9 (top)
- hol2 (top)
- ho10 (top)

basically a "hole" minus the bottom "ho10"

may be make use of "hol1" for these?

So only axlehol7 drops out of the pattern

[Gerald Lasser]

Fitting axleholes into parts is something we should address IMHO.

Mostly they are embedded into either a ndis or a ring primitive. Once a Prim-Substitution is done, gaps appear, as the axleholes themself are mostly just surfaces. So it looks that seperate surfaces to embed an axlehole might seem the better solution to avoid these gaps.

I noticed that

axl2ho10
axl5ho10

are using 1-8chrd prims on their "thicker" parts of the axlehole to better fit to the prims that surrond them.

Additionally, the axl2hole is using a 1-4ering to ocver the thinner side.

As this is not uniform and to make the axleholes interchangable AND avoid gaps, there might be two solutions:

a) Not using chrd and ering in the axle prims
Let the user of the axlehole decide hwo to embed it, so the user can choose if it will be embedded in a ndis/ring and add the necessary 4-4ering themselves, or use triangles/quads

b) using chrd/ering in the axlehole prims
This would imply that an Axlehole is always embedded in a ndis/ring
Embedding it in surfaces would lead to overapping surfaces


I would prefer a).

Nevertheless, I think in the primref we sould give a "best-practice" on how to use those in the end

[Philippe Hurbain]

I tend to prefer a) too