topic Re: pattern becoms inverted :-( in 3D Part & Assembly Design

pattern becoms inverted :-(

jlynch — Wed, 15 Jan 2020 12:05:14 GMT

Hi All,

I need to create a pattern like this...

I've built it about 10 differnt ways, but it seems to always fail on the inflection point here

I'm sure it something to do with normals being flipped for the curve as it goes over the apex. (start points on curve flipping from botttom to top - if you copy and paste the middle curve from the first group in the patter and conpare the direction start point to one from a failed group you can see what I'm talking about )

basically what I've done is create a spline, place a point at a real distance of say 5mm from start, another at 10mm and another at 15mm. (I tried patterning this 3 times to get 3 points at 5, 10 and 15 for construction but then Creo was complaining about patterning that)

I then constcted 3 curves that were normal to the spline and through each point

then a new sketch over that to add rounds

then I pattern that construction and final sketch to give me the geom I need.

I did maybe 10 variations of this, diferent ways of constructing the same geom but they al fail on inflection where the arcs on the last sketch jump to the opposite end

I've tried also constructing offsets and using edge and trimming to get the underlying geom to add full rounds afterwards but it just fails at inflection again.

If I unpattern it, and go to where it fails, a simple sketch update references solves it..

it's very annoying.. 😖

Creo 5 part file attached for anyone willing to take a look.

any help appreciated

Thanks

James

Re: pattern becoms inverted :-(

DavidBigelow — Wed, 15 Jan 2020 17:01:01 GMT

Yea - this is a bit of a frustration. Often I find that if you keep things more local you can get better results. I noticed your patterned used three reference offsets... this should be done by one. I used a perimeter dimension to simulate the offset distances that you were using along the point on curve. You get the same distances for the resulting points from the center of the reference - then built from center point of the pattern out.

Another problem is the tangency of the arcs... they will never be able to be perfectly tangent to the normals under a spline trajectory. To "hack" this - use a spline and tangency to mimic the arc tangency - this assumes that your goal is to have the tangency on the outer spline offsets -- which are always varying their gap distance along the trajectory.

IF you do not need the tangency for the arcs to be on the spline - then the length of the normal curves would have to be allowed to extend beyond the offset splines (in or out depending on the inflection).

This is what my updated version looks like.

Attached is the file -- hope this helps.

Dave

Re: pattern becoms inverted :-(

jlynch — Wed, 15 Jan 2020 17:36:19 GMT

Very Cool Dave and thanks a lot for the time and effort!

love the simplicity of just driving the sketch through some points.

The rad tangency is easy solve with a reference pattern and a round feature, I only put it in the sketch after about my 5th attempt to join the normal lines.

Thanks again 😉

James

Re: pattern becoms inverted :-(

pausob — Thu, 16 Jan 2020 04:15:18 GMT

Out of curiosity, how did you even get your step #2 to work? If I try to sketch a line and use for the sketch references: previously sketched spline and a datum point that was placed on the same spline - then I'm unable to draw a line that is on the point and perpendicular to the spline... Upon trying to impose a perpendicular constraint, the system says "Invalid selection; try again". Maybe that's a Creo 4.0 limitation?

Re: pattern becoms inverted :-(

DavidBigelow — Thu, 16 Jan 2020 04:21:55 GMT

Try to create a datum plane thru a point on the spline AND perpendicular (Normal) to the spline as the creation constraints.... then you have a stable normal reference where ever the point moves along the spline.

Then use the datum plane as your vertical sketch reference.

Does that help?

Dave

Re: pattern becoms inverted :-(

pausob — Thu, 16 Jan 2020 04:38:06 GMT

Well, yes that works. Sure seems like an unnecessary and esoteric work-around 🙂

In fact, if I try it with a curve comprised of lines and arcs, (i.e. not a spline), then

are no issues and I don't have to construct that "extra" datum plane...

Re: pattern becoms inverted :-(

DavidBigelow — Thu, 16 Jan 2020 19:30:41 GMT

Agreed - but this was a spline... 🙂