Twisted sweep

[vitulin]

Hi,

i hope you can help me with the following problem. You would have a oil dipstick, which would be general 3D curve spine. And you would need to create twisted section on it. In proe, you would do it the way, that you would simply draw a line in the sketch and constrain it with angle. And than you would do it as variable parameter for example angle = sin t...where t would be 0 to 1 of the curve. Do you have any suggestion how to do it in UG?

Thanks
Vit

 

[hudson888]

Vit,

Sorry I couldn't get back to you sooner with and example so it looks like John has done most of the work already.

I may be ignorant of dipsticks but if you check my examples I was of the opinion that it should probably be twisted along the axis of the metal strip. Therefore I varied John's example in a couple of ways.

The second way that I varied the construction was to describe the twist law using a bridge curve rather than a sketch. I did this for no other reason than to try it out really proving you could be something a bit different.

Lastly in the attached is a second example of a bent dipstick. This is loosely based on the file which you attached. In that file you used a line which was NOT normal to the plane of the sketched section that you're sweeping. To me this was not likely to represent a desirable solution, and if indeed you need to follow such a line you would simple re-orient the other geometry to align with that as a new sweep guide and twist axis, the two in this case being the same thing.

Whereas in my example I have provided a curved dipstick with a twist the arc being constructed normal to the axis of rotation which just uses a line.

John,

If I may tack onto the end of the post for Vit here. As usual your construction is superior to what I briefly described earlier and I thought it a good example. I tell you what though it is not easy to figure out how a sweep has been built. The dialog when you create it presents one set of inputs and the edit parameters has something completely different. I hope in upcoming versions that this will be addressed sooner rather than later. I find sweep a very powerful but perhaps underutilized construction method, that is frequently poorly understood and probably languishes thus for being somewhat difficult to understand more for the presentation than the actual content.

Cheers

Hudson

 

[vitulin]

Hudson,

thank you for the examples... Unfortunately, its not exactly what I wanted...I dont have a problem with using a curve to control the angle along the curve...the only thing is, if for example height of 100mm means 100degrees...but it seems to be...the problem what I had was, that I tried to control this by the expression, which doesnt seem to work...it seemed like the parameter t is still constant rather than changing with the position on the spine...so i would like to see, how you have to set up the set of expressions for the t parameter to work...i think, that UG basicaly wants t and ft(t)

 

[hudson888]

Vit,

I think it is just a vertical and horizontal scale height in the Z axis maps the range over which the twist occurs so that the straight portions of the law curve induce no twist and you can use two arcs, a bridge curve (as I did) or even a straight line in between to control how tightly the twist occurs in the section of sweep directly opposite. On the horizontal scale in the Y axis 1mm of arc length is gives you one degree of rotation.

You can experiment with different law curves if you like just to help you to better understand the results.

Hope this helps

Cheers

Hudson

 

[vitulin]

Hudson,

and could you help me with the equations?

Vit