Hi,
V5
I search youtube surely someone has done the basic of how to model a nut.
BUT NO.
some japanese thing with zany music.
Vimeo has 929 tutorials but no search facility that I can see.
How does one find the one on doing a nut ?
Having used hexagon to get solid with correct face to face dim, I just want to see how one gets the curved corners, do they arc or go straight down their slopes, and what decides the shape there.
Its the sort of thing that will take an hour to draw a simple nut, time lost on trying to find a tutorial.
yes there is a library of fasteners out there BUT last time I used it it polluted my file.
??
just BooleanIntersect with a mirrored truncated cone ? (I build the it as revolve ā¦)
if you want to build the thread as well - only draw half turn of one pitch
(_helix _sweep1 or sweep2) - then rotate, arrayLinear
building a thread with all turns in one go might give troubleā¦ M10.3dm (3.6 MB)
I donāt believe the āchamferā size is specified anywhere. I also use @Tom_P 's cone/chamfer method when I teach making bolts/nuts, I use a cone angle of about 30Ā°. The cone should just ānickā the straight edge at its midpoint to produce a tiny chamfer, the rest of the shape will just happen as a result. This shape is actually the result of nuts being classically manufactured using automatic screw machines, the cone being what is cut away with the chamfering tool as the nut rotates.
That being said, IRL, these things can be all over the place. Here are three M10ās I picked out of my hardware collection this morning:
The one on the left appears to be classically machined, the chamfer is clean and the edges just broken. On the one in the middle the chamfer is more exaggerated and slightly rounded - itās no longer a simple cone revolve. The one on the right can no longer really be called a chamfer - the shape canāt be made on a screw machine, itās totally rounded and looks more likely to have been stamped into that shape.
Yep, the chamfer is a bit big on those, doesnāt leave much flat bearing surface to compress the load - which is kinda what you want a nut forā¦ but they look nice.
Hi,Tom, thanks.
No not going to buy result.
Its on the end of an item with a flange attached to it and a thread at the rear for screwing into something else.
It appears that the user decides how much to go near the flat edge and the angle again take measurements of the item being replicated and adjust angle to match.
The vids are all in foregin lingo, as are the texts on the rhino interface.
Not one english speaker has shown making a nut.
However from Toms initial diagram the method is clear. even a triangle spun round as solid will suffice.
The flat top is generated by the side, not the other way round, tail wags the dog !
ikr. they used to not even make helix threads, back in the day
theyāre files are pretty good for mockup / rapid prototyping. but i usually have to purge the block instances, or repair or recreate their geometries.
well with anything, IRL, as @Helvetosaur mentioned, the machine process dictates the shape.
And post processes as well.
However, the design intent is supposed to be what dictates what processes are compatible to create the aforementioned design intent.
The design could follow certain industry standards as @Tom_P mentioned, but still the design intent is upto the designer.
Then the actual IRL outcome can dictate the cost of said design or manufacturer talent etc.
You decide the shape as the designer. If you want industry standard specs then you might need a machinist handbook and/or machine-shop-101 trade-secrets etc.
Indeed, something as simple as a nut could easily take hours and hours to model, depending on the details.
Yes, if you download something from a library somewhere, it will just be some model that some particular modeler modeled for some reason, maybe they were paid by the hour or something ā and maybe they didnāt follow very good modeling techniques and their model is full of bad or inefficient geometries including block instances etc.
If you really wanted to take care, and the bolt was heavily loaded, you would check with some form of FEA/Stress Simulation.
In reality, if it wasnāt doing anything special, I would probably just turn a generic 0.25 mm chamfer or radius on it (the hex) to start with, and see what it makes. You can emulate this in Rhino without RhinoCAM.
There are all sorts of different ways, rounded edges, milled chamfers, or standard turning. Just depends what your requirement is.
Honestly I havenāt researched this for a few years, so I just did a quick review, and yes Iād agree the results of videos available on this mater are really bad
I made a video about this process for a job interview several yrs ago, but the guy never got back to me, so I deleted the video. And it was more about the threads. I reviewed my video from my archives last night, and it was pretty bad lol
Most of the videos Iām seeing, it looks like the modelers donāt even know what 60 degree threads are
Most likely theyāve never had machine shop 101. Not sure if I should create a new video
Hi,
the nut I need to make isnāt screwed onto a thread or bolt, in fact I suppose it could be called a bolt, it has a 1 inch dia thread (1.06 in fact) aft of it and a hole through its middle and through the thread !
It wont be in any manuals, I just need to recreate it for a visual recreation of it in metal by someone and the rest of the item.
Please CAD CAM CNC do us all a favour, show how Rhino can make a nut matching an existing one.
I reckon the steps are, with the non standard 80 year old nut in ones hand ! :-
measure across flats, (more accurate with a vernier, use hexagon tool and scale result to the flats dim.
measure height of the highest part of the side, make a solid that height using hexagon curve.
measure the shortest height at a corner and plot on the solid with dim and perhaps a point.
measure the dia of the flat top and create disc to match.
take line from edge of disc through point in step 3. then other clicks to make a profile for a cutting solid.
rotate this using surface comand and cap any planare holes.
boolean difference this from the hexagon solid.
both sides if a normal nut.
establish thread data and create thread then boolean difference the thread.
I would consider doing this as I would machine it. As long as I had no particular surface requirements.
Rhino (CNC/Machining)
Make a revolved profile (turn bolt/nut head).
Trim/boolean out the hole in the shape (Centre drill and drill for nut internal threading or grinding).
Boolean away the revolution, using an inverted hex shape (mill flats for hex head)
If the profile isnāt turned smoothly, then fillet/chamfer the edges (chamfer milling)
If the hex isnt milled smoothly, refinish to create chamfers (fillet hex edges)
From left to right, the item:
High-Strength Class 10.9 Steel Hex Head Screw, M8 x 1.25 mm Thread, 70 mm Long, Fully Threaded
From McMaster.
Then a series of models images produced using an approximate profile of that screw, showing the process as you go.
Honestly, this thread got me pondering alot about this subject lately which also just so happens to present a great opportunity to reflect on the subject of parametrics.
It would seem that this entanglement is more common than one would first expect. I mean who doesnāt want to be able to easily make threads haha.
Ahh yes the proverbial āvisualā that gives 3D models the amazing look an feel, that brings imagination to light.
This reminds me of all those models I did in college along time ago, and put real true threads on them when no one else would. Back in those days most CADās had a silly annotation texture wrap for āfakeā threads lol.
Well thereās a handful of ways to do this, so it would depend on the requisites, but I suppose we can start with the first way, and just go through them all ā the last of which being the super technical route.
Yes that sounds about right.
Later, in terms of āthreadsā, there will be a thing called āmajorā and āminorā diameters ā super tech route.
Um, Iām not sure, but yes in the long run if enough degrees of freedom can be reduced or rather ādefinedā by a parameter, then the outcome will be a nice parametric nut or bolt or screw etc.
Iām not certain, but yes the more measurements the better ā meanwhile, some measurements can be derived from the knowledge of others; this is why most āblue printsā donāt have to specify every single dimension of a particular component within a particular schematic.
Sounds about right.
Sounds familiar, but the details can be adjusted if anything needs refinements.
Also sounds possible.
True, but whether you ācapā or ābooleanā etc. is just a few of about a dozen ways to get the final result. It all depends on the modelers own tendencies and desires or techniques etc.
Indeed. Whatever gets the job done really. Different workflows may require different setups to work, but getting the final result of which should be a āclosedā polysurface or something, as the end goal.
And if the end goal is simply a āvisual recreationā then the super technical parts arenāt as necessary.
From a machinist standpoint the threads specifications can be complicated if you have to pull out those thread gauge kits with those wires and charts and suchā¦
Yes, I agree. I think itās definitely a great approach to think of all 3D model features as one where how it gets created in real life.
Even in real life it can be surprising how a particular model might have to be revised in order for the real life process to actually succeed in manufacturing and mass production.
You never know when youāll need a certain feature, just for the sake of obtaining the actual design intended feature.
For example, when machining plastics, you might actually need a 0.005" radius to machine a āsharpā edge with āno burrsā. And you might have to āpeelā the toolpath in certain directions etc., to deal with burrs.
I might put some time into this. Iām sure thereās other members whoād also do a great job in this regard. Iād be happy to do the honors of tackling this exercise. Although, Iām not sure what all the parameters are upfront. I might have to just take it one step at a time and do revisions. And possibly create anew thread at some point.
Iāll most likely try to use Grasshopper for this, ultimately, so as to get parametric control. But I still have lots to learn in GH.
In the meantime, Iāll try reviewing technology like:
A couple years ago (when I was first starting out with Rhino) I needed to make a part with an uncommon 40 tpi pitch. (itās actually fairly standard in the optical field but in generic hardware it isnāt).
This video was very helpful then - and guided me towards a working part, particularly the approach for building arbitrary threads.
No narration but pretty easy to follow what is going on.
Back in those days most CADās had a silly annotation texture wrap for āfakeā threads lol.
