Converting sheets to an arrayable solid 1

[SiW979]

Hello all

I've been struggling converting a bunch of closed sheets to a solid that is arrayable around a cylinder, I'm using NX4 and in the past I have used patch to join sheets to a solid body and also extract to create an associative body from a swept feature in order to array the feature around a cylinder etc. However in the model of the tyre I'm creating I can't seem to make the sheets in to a solid by sewing, so that I can then unite to the tyre and create the array. I have attached the model to this thread for those who might be kind enough to put me out of my misery. Any help will, as always, be greatly appreciated.

Best regards

Simon
JCB

Best regards

Simon

 

[hudson888]

Simon, QED...

I think your sewn sheets may have worked but I saw a much easier way and went with it. For the array you group the features first then you can array them. Have a close look and you'll see that I united after. In NX-5 you don't need to know this.

Is there any point at all to the sheets? I didn't want to rebuild your associative sketches so I left the cylinder as it was. However you could just as easily make the whole thing using solids from this example. I did notice that you scaled it way down so you may have oversimplified. Let us know if you're still stuck.

Regards

Hudson

 

[hudson888]

The attach didn't work Sorry!

It turns out that you have to upload, then hit attach, before you submit.

Good Luck

Hudson

 

[vitulin]

Hi Simon,

Please see the attached model. I hope it helps you.

Regards
Vit

 

[hudson888]

Simon,

For interests sake I had a go at the sewn version. It seems like it doesn't cope with trimmed sheets too well so in this case trim body provides the substitute method.

Good Luck anyway

Hudson

 

[SiW979]

Hudson, Vit

Thank you for your help, it appears as though my sewing method did work however it didn't change colour and show as being solid which was cufusing me.

So I've managed to create the array now and I've gone to start applying some draft to the tread but even though when I check apply draft to all instances, it only applies it to one, the blends however are applied to all instances as expected. I've attached the model again, do you have any ideas please?

Thanks again guys

Simon

Best regards

Simon

 

[hudson888]

Add draft to it before the instance

Best Regards

Hudson

 

[hudson888]

Or simplify the whole thing by using pattern face.

Sweet

Hudson

 

[hudson888]

Vit,

I'm not sure about the extracted body, though I followed your reasoning I expect we were in a bit of a race to do the same thing at the same time as our thinking was otherwise quite similar. After we've all had a couple of goes I think the pattern face is probably the neatest result.

Regards

Hudson

 

[SiW979]

Hudson

Pattern face is without question the neatest method so far, although I was a little disapointed that it wouldn't work with the 3 general pads I've added to the model, nor does cirular array, so I guess I'll need to use the extrude and extract method.

Once again thanks for all your help.

Best regards

Simon

 

[vitulin]

As for the extracted body...as you all know, when you want to crate an array of solids, UG will not allow you to do that, unless it can add it to some body...but if you do a little trick and either extract the solid or group it, you can then create an array without the need of having the unite or subtract feature...

 

[hudson888]

Vit,

The other trick pre-NX5 was to just add the feature/s to a group which then allows you to array them without even requiring a manifold boolean. Tha was in one of my attempts .

Simon I'll have a look at your latest offering and post separately.

Hudson

 

[hudson888]

Simon,

In this case with three separate general pads you'll find that Pattern Face is based on the use of regions with boundaries and you can only define on such region to pattern at a time. So you see you would need three pattern face commands to complete the task.

The generals pads that you have used are a far more discrete and self contained feature that the larger treads which consisted of several elements not collected as closely in time-stamp order. For those reasons and for efficacy I chose to group the three general pads and then create an array based upon that grouped feature. It just goes to show there are benefits of either method and you could chose either without losing any sleep over it.

I will attach a Winrar archive, for those who need that software it is available as shareware for trial at no cost. I found a 6+Mb file getting a bit large so that the upload was failing.

Best Regards

Hudson

 

[hudson888]

Simon,

Before I finish on the topic I have made you a model here that I might better recommend in terms of showing the methods I would prefer to employ. I also found that the base radii of the outboard general pads were not so good in the course of trying to find out why pattern face failed to work. It seemed there were some persistent tiny objects created with that general pad perhaps hinting that other more painstaking methods may yield a better result. Although I'd have to say the you'd need to be quite keen to bother.

Also check out the second draft, you don't really need to create a datum plane each time. And the way I created the extrude to rebate the top of the main treads can be done by a different perhaps more direct method.

Best Regards

Hudson

 

[SiW979]

Hudson

Thank you for your time and explantions, they have been a great help.

Best regards

Simon

 

[vitulin]

Hudson,

sometimes I use just the grouping too. But when you for example have linked body, then just grouping does not work. Thats why I sometimes use extract body and then I group it with the following features.

Regards
Vit

 

[hudson888]

Vit,

I just checked using NX-4.0.4.2 and was able to create a group (feature set) containing a single linked body and then use that feature set in a non-manifold circular array. I'm not sure in that case if there is any need for the extracted body. Have another try without it if you like and see whether it works for you too.

I would avoid having extra solids in a file wherever possible as was the case for Simon's example. Otherwise you have always to take care that they find their way into neither reference sets nor weight calculations. If you're careful you can avoid it of course. Like you I just wanted to cover all the bases and show what I would recommend.

Best Regards

Hudson

 

[vitulin]

Hudson,

I maybe didnt express it exactly...when I link the body, add it to main solid and then blend or something...and then group it and try to create array, it doesnt work..it says, that it cannot create tool...

Vit

 

[hudson888]

Oh yes now I get you. It doesn't always happen, and even after extracting a body in the tyre example it didn't get me very far.

If you're going to be able to unite or subtract your body using another body then you may be able to do so as Simon did in one of his attempts, by creating a circular array of the unite. Which was I suspect where the post started because in trying to have the drafts and blends applied after the unite and before the array, this "unable to create tool body" error may keep coming back. That's where the pattern face comes in because if you can use it then it seems to me that you can simply unite the linked body without all these problems.

Now the error "unable to create tool body" isn't restricted to linked bodies in this particular case I noticed it in Simon's original part as well even without a linked body.
So what I'm wondering is whether your idea that creating the extracted body is actually guaranteed to work on all occasions.

To be fair neither have I absolutely exhaustively tested the pattern face method. I managed to have it refuse to work with some of the general pads for what I think were different reasons. From what I can see it does not at least have trouble with linked bodies.

Best regards

Hudson

 

[SiW979]

Gents

I'm happy to report I have completed the model (image attached) and once again I would like to thank you for your help, the forum has once again been worth it's weight in gold.

Best regards

Simon