Issues with generating Threads

Cadman

I have an issue with generating (or more specifically, re-generating) threads once you make a change to the Pitch.

I created my Helix, created a triangle to define the thread perpendicular to the end point of the Helix, as shown in the latest tutorial, and then performed the sweep. The thread generated just fine.

However, when I went back and changed the pitch of the Helix, which changed the end point, which made the sketch that defined the thread no longer perpendicular to the Helix and the thread (sweep) did not re-generate.

Any ideas on how to prevent this? OR Am I missing something?

Thanx,
Randy

matthew_menard

Any suggestions for making threads mesh between parts?  This has been a pain for me to do in the past in Soldiworks, but I figured since I'm making all of my parts in one workspace it should be easier.  I was playing around with some shared sketches but was still having trouble getting the threads to mesh.  Should I try and share a helix between threads as well?  I'm trying to model the threads using the parallel plane to axis method, not the normal to the helix method, if that helps.  Thanks for any tips.

_Ðave_

you need to create the triangle on the starting end of the helix, the starting end will not change when the pitch changes. unfortunately that is where you likely have the head. Maybe someone from Onshape can tell us how to determine/change the starting end of a helix when creating it.

coleman

@randy_11 simple solution: 

You need to sketch your triangle on a curve point plane.  This will ensure your cutting tool is always tangent to the helix start point.  

Check this thread for more info on threads and curve point planes. 

andrew_troup

coleman said:

@randy_11 simple solution: 

You need to sketch your triangle on a curve point plane.  This will ensure your cutting tool is always tangent to the helix start point.  

Check this thread for more info on threads and curve point planes. 

Everyone should be aware that, although this is a common practice in solid modelling, it does not strictly comply with the geometrical underpinnings of a proper (ie standardised) threadform. The difference will not however be appreciable or even detectable in practice except for very coarse threads, particularly multi-start ones, but such threads do some up at some point in most people's careers, which is why I raise the point.

This topic came up recently in another thread (see *1), and it's a very popular misconception that the straight-sided triangular master profile should lie on a plane whose normal is tangent to the helix. Strictly speaking, the profile should lie on a plane parallel to the axis of revolution of the thread.
Even people who know a lot about precision thread production sometimes get this wrong, perhaps partly because when grinding such threads (which is done for either hardened workpieces or highly accurate threadforms or both), the grinding wheel is tilted to the helix angle, and when inspecting finished threads non-destructively, the optical comparators are similarly tilted.

However this is out of necessity, to clear the approach and recess flanks: the profile being crushed into the grinding wheel, or loaded to the comparator, should be corrected so that it is no longer straight-sided.

*1: ON EDIT: it wasn't a discussion thread, it was in Tech Tip: Creating a Thread in 60 Seconds

traveler_hauptman

@andrew_troup  Great point! 

coleman

andrew_troup said:

coleman said:

@randy_11 simple solution: 

You need to sketch your triangle on a curve point plane.  This will ensure your cutting tool is always tangent to the helix start point.  

Check this thread for more info on threads and curve point planes. 

You are 100% correct.  However, the purpose of modeling threads is for demonstration (or marketing materials).  CAM software doesn't need modeled threads and thread callout on a part print is simple without actually having the thread modeled on the part.  

I wonder how this will effect the 3d printing folks.  

andrew_troup

Actually I was wrong to say above that the triangle for a thread should always be on a plane parallel to the axis of the thread.

For high helix Acme threads, Worm threads, and the like,the profile conventionally lies on a plane normal to the helix. The distinction is unimportant for most 3D modelling, but I thought I should correct the misinformation because people who read these pages might end up doing other stuff, and it matters a lot for thread grinding and suchlike, especially for (say) hardened worms for power transmission. 

kevin_quigley

Threads and modelling off are a mystical black art! My rule of thumb is this. If it is a machined thread don't model it...add as note to drawing. If it is a moulded thread, model it. As 99% of the threads we do are for moulded parts we tend towards modelling, especially as those threads are rarely "standard". Again most of these are created using the thread profile on a plane normal to helix.

matthew_menard

Any suggestions for making threads mesh between parts?  This has been a pain for me to do in the past in Soldiworks, but I figured since I'm making all of my parts in one workspace it should be easier.  I was playing around with some shared sketches but was still having trouble getting the threads to mesh.  Should I try and share a helix between threads as well?  I'm trying to model the threads using the parallel plane to axis method, not the normal to the helix method, if that helps.  Thanks for any tips.