Arc Pattern - Chord Length

[Dale_Rosema]

I am looking for a way to make a radial pattern given a horizontal dimension from the circular centerline and a radial dimension from a point on that centerline and the spacing is a chord length.

The first hole is 6" horizontal from the center line.

The beginning of the radius in 16" below the horizontal centerline.

The radial dimension between holes is a chord length of 4".

To complete this I backed into it by calculating the angular spacing given the radius and the chord length. Is there an easier way to pattern this than manually calculating the angle?

[Patriot_1776]

'Sup Dale!

Should be a way to "math" it using the diameter and perimeter of that diameter, say, as parameters.  I'm swamped or I'd look into it for ya.

[pausob]

To be honest, I don't quite get the dimension schema for your pattern; lol, the only thing I think I understand is that the patterned hole centers are 4" apart.

But I was going to suggest that instead of "manual calculation", you create a sketch which computed the necessary axial-pattern angle, such as this:

for a given radius (R=9.223" in this case), and with the chord spacing of 4", the necessary angle is displayed in the reference dimension of 25.049° - and that reference dimension can be used in the relations that drive the subsequent axial pattern...

[Dale_Rosema]

Let me break down into a couple of things:

Is  there a way to simply define the location of the first hole. It is a combination of X, Y and R, Θ.

I have a 6" horizontal offset from the vertical centerline. I also have a radial dimension from a point a that same centerline.

Without creating sketch geometry, defining a point and then putting the hole on that point, I could not find a simple way to define the hole location. After setting the horizontal dimension, I would click on the center point and Creo would default to a vertical dimension and not radial. If I chose a radial dimension first, Creo would not let me chose the vertical centerline for a second dimension.

The second issue was trying to determine the possibility of defining a pattern by chord length instead of angle. The holes are the same chord length apart, the angle dimension changes as the radius get larger from row to row. Can you define a pattern by chord length?

Yes this all can be done by sketches or math, but I am just trying to figure out the possibilities of the software. Why do it the long way if there is a shorter method that could be utilized of which I may not be aware.

[StephenW]

If you sketch out what you want your dimensioning scheme to look like, it might help get a better answer.

I think you are going to have to make it a sketch to get the dimensions you want, but that's based on what I think you are after. I not sure if i am getting lost in the description or not.

On a good note, I think you can get to where you want to be pretty easily with a sketch and then a hole, then pattern

[Dale_Rosema]

It's not that I cannot get it made in Creo. I just have to make a bunch of sketches.

The first question. Can I locate a 2" diameter hole on a surface that is 6" right of center on a 27.4838" radius (sorry, should have edited the dimension to something easy but was quick sketching this out in Draftsight)? From what I was trying over the weekend, I can either go X,Y  or R, Θ, but not X, R. If I put the 6" dimension in, when I try and made the radius dimension (27.4838") it gives me a vertical dimension instead. Which puts the hole slightly above where it is supposed to be.

Once that first hole is established, Can I pattern that hole with a 4" chord length? (I can calculate the degrees given the radius, but if I add a 2nd row of holes at a radius of 31.4838, I have to recalculate the degrees when I would just like to pattern at a chord length.)

[StephenW]

I'm on creo 8.

Use the "sketched" hole and only sketch points for the holes. I used "construction geometry" in that sketch to get the dimensions you need.

Then pattern, selecting only the 

[joojaa]

Which one of these are you after, if any?

A) uses a partial construction arc, and arc angle measurement. You can make a arc angle measurement by clicking arc/curve endpoints then the curve itself. Arc angles tha a arc angle symbol on top.

B) make a point measure distance from point to point then edge to point.

The resolts are obviously different

[Dale_Rosema]

I am looking for "B".

I can use points, but just wasn't sure if there was a way to drop a hole on the surface and drag the (2) handles to be horizontal and radial.

[joojaa]

No, the handles are just for some use cases. not all possible combinations. Obviously there is a million dimension schemes you could use instead, all those generic schemes are sketcher solutions.

[Dale_Rosema]

The other question then is chord length patterns? Go or NoGo.

[StephenW]

Neither of those dimesions are chord length.. I can't think of a way to get chord length without a sketched line.

The sketch in mine has chord length but it 's really hard to see.Chord length would be point to point, not datum to point or arc length.

[joojaa]

Sure. However chord length from point to next point is not really possible in generic sense. However, since it is a circle you could calculate the arc length to chord length conversion or alternatively chord length to angle in a relation. (Or you can draw one chord length and measure angle it in the sketch with ref and use that angle to drive the pattern) No real solution for any other shape.

[Dale_Rosema]

That's what I did. I had a conversion from chord length and radius to angle and plugged that into the pattern. It is just that you have to calculate it for every row in the array. since to maintain the same chord length with a larger radius means a smaller angle.

[joojaa]

You can write the relation in a dimension pattern, then it works regardless

[StephenW]

Thats why I liked the sketch point in the hole feature. Everything is in one feature. No relations and can be patterned without thought.

[joojaa]

Yes but if its a highly parametric model, lets say it goes to some sim optimization. Then it is not a solution, who would like to mindlessly draw 1000 variations of the pattern.

[joojaa]

Unless i am mistaken a relation in a dimension pattern has a form:

dist = 6

dia = 2*27.4838

raddec=10

memb_v =2*asin(dist/(dia-2*raddec*idx2)))*idx1

and sanity check

seems to work.

[pausob]

wait... what?  trying your relations with a 3 x 2 pattern:

 --> 

Clearly I'm not able to get the results you are showing.

Your relation has errors - too many right brackets?

What dimension are you controlling with the relation shown?

Is there a relation for the dimension in the 2nd direction of the pattern?

@joojaa , please elaborate -  it would be great to be able to learn from your example.

[joojaa]

Oh sorry shouldn't comment things on the fly (so i tried to convert my relation to a bit more readable form but fumbled). So i am controlling angle and then the radius. You are almost there.

So lets assume you have a the following scheme (dimensions not super important):

Then doing a dimension pattern on d1 in one dimension and d3 on the other assuming you want a distance of say 6 units (make this a parameter then you can easily change it as you need, but here i hardcode it into the relation to save making more images or longer explanation)

Then the relation is (im going inwards but you can go outwards too if it makes more sense):

dist = 6
memb_v =2*asin(dist/(2*d3-2*10*idx2))*idx1

After this i edit the first pattern member to have a angle of 0 (fits item is not controlled by expression) and i Get:

OK so one deficiency for ease of following. I have hardcoded also the second direction increment but that is a parameter so i could change the number 10 to relation after the fact (in my case d6 but check yours) You could also do the same with the dist so then its fully editable.

[StephenW]

I definitely got confused on what you were needing. LOL, Oh well, back to the looney bin!

[joojaa]

Yeah the original post was confusing that why i needed clarification. I think this is now what OP needs. Not sure though.

[Dale_Rosema]

Each row does no necessarily have the same number of holes.

[tbraxton]

"From what I was trying over the weekend, I can either go X,Y  or R, Θ, but not X, R."

If you want to use X and R as the parameters driving the pattern, then I am pretty sure you will need to use relations as the axis pattern will not take those inputs directly AFAIK. You can express the chord length and the X,Y points of circle as a function of R. You have values for X and R and can use those to find theta which you have already done. If you then also define the chord length using X&R in a relation, then this can be reused in the model to locate all members of the pattern lying in the base radius for hole centers.

Will the pattern be confined to a single quadrant (90 deg) or do the holes need to be spaced around 360 deg of included arc? An axial pattern is attractive for ease of creation, but a relation driven dimension pattern may be more flexible if that is needed.

If you look back at this post and the new functionality added to axial patterns in Creo 10 it may be useful. I was able to implement that geometry (pre Creo 10) using a pattern of a pattern. PTC claims it can be done in a single pattern in Creo 10+.

https://community.ptc.com/t5/3D-Part-Assembly-Design/Reduce-to-practice-a-variable-axial-pattern/m-p/750566