topic Re: Bug. 10-32 UNF modeled thread! in Fusion Support Forum

Bug. 10-32 UNF modeled thread!

HughesTooling — Tue, 17 Mar 2020 16:17:42 GMT

If you set the Class to 2A the thread fails, works OK with 3A. Also if I set the thread to UNC it models correctly with 2A and 3A. Model is attached just edit the thread and change the class.

Mark

Re: Bug. 10-32 UNF modeled thread!

MR-R — Sun, 01 Nov 2020 23:19:20 GMT

I was having trouble modeling a simple 10-32 UNF thread and came across this thread.

Does anyone know if there is a fix for the error? I'm wanting to create a piece of threaded rod so that I can create an assemble with two other parts, both of which were able to be modeled with 10-32 UNF internal threads.

I know that changing to 3A as mentioned below is one way around the error, but my internal threads are class 2.

Re: Bug. 10-32 UNF modeled thread!

HughesTooling — Mon, 02 Nov 2020 09:57:54 GMT

@jeff_strater Looks like this post was missed!

Looking through the XML file it looks like a typo for the pitch diameter, it's set to 0.6673. Probably should be 0.1673

Re: Bug. 10-32 UNF modeled thread!

HughesTooling — Mon, 02 Nov 2020 10:13:24 GMT

@MR-R I've attached the UNF XML file with the pitch diameter set to 0.1673 and it seems to work OK. If you do a search for ANSIUnifiedScrewThreads you should be able to find the directory with all the thread XML files. Just be aware when Fusion updates you'll probably need to manually update the file again.

Re: Bug. 10-32 UNF modeled thread!

TrippyLighting — Mon, 02 Nov 2020 10:54:08 GMT

@jodom4 can we get this fixed?

Looks like extremely low hanging fruit.

Re: Bug. 10-32 UNF modeled thread!

TheCADWhisperer — Mon, 02 Nov 2020 11:04:53 GMT

@TrippyLighting
Looks like extremely low hanging fruit.

Picked, dewormed, washed, peeled and diced by @HughesTooling . Ready to be baked in the pie.

Re: Bug. 10-32 UNF modeled thread!

rhdfmail — Mon, 02 Nov 2020 15:10:31 GMT

Noticed somethin similar, but for metric threads when using hole-command.

A tapped M5 hole without the tread modelled, when measured its 4,234mm. wich of course is based of some of the dimetions of the actual thread. Would be nice if the hole would be the recomended drillsize

Re: Bug. 10-32 UNF modeled thread!

seth.madore — Mon, 02 Nov 2020 16:24:42 GMT

@rhdfmail wrote:

..Would be nice if the hole would be the recomended drillsize

Well, that might become a bit of a Pandora's box right there. What's the desired percentage of thread? What's the material? Are we using a form or cut tap?

Re: Bug. 10-32 UNF modeled thread!

rhdfmail — Mon, 02 Nov 2020 16:51:43 GMT

Yep..

But following scenario is possible.
Someone designs a part and uses the hole-command for all threaded holes.
I program the part and sees M5-thread on the drawing and doesn't measure the holes and instead imports my "drill & threadmill for M5" template.

Here 2 things can occur.
#1 I have seen by threaded holes in solids before where the hole for a M5 can be anything between 4 and 5 mm and tuned my drill-operation to just use 4,2mm holes (so I notice when holes that is supposed to be included isn't)
Here I then have to either change the holes in the solid or adjust the drilloperation AND the threadmilling

#2 I didn' do that and instead specified 4 -4,5mm
This means I will drill a correct hole at the right spot, but my threadmill-op will use the 4,234mm hole and apply its diameter offset from that (because someone thougt that approach for treadmilling was a clever idea)

This means I will get a bit too big threads (not much but noticable)

If I would design every part myself that would not be a problem. Just adjust my template to correct values. But 99% of what I run is designed by someone else.

Re: Bug. 10-32 UNF modeled thread!

jeff_strater — Mon, 02 Nov 2020 19:10:32 GMT

thanks for reporting this. Created FUS-74291 to track this bug.

Re: Bug. 10-32 UNF modeled thread!

jeff_strater — Thu, 05 Nov 2020 17:17:01 GMT

This problem has been fixed in our development version. I'm not sure exactly what update this will appear in, but it should be relatively soon. Thanks again for reporting the bad value in this thread spec.