Parallel Arcs Sketcher to Part Design Problem

[Gray Area]

Here is the latest iteration of my part.

When it is completed, it will be an accurate drawing of the “gas tank support” of a C2 Corvette. I still have to place the mounting bolt holes, and input the actual measured dimensions. But, as is, this is roughly what the part looks like. As it is a parametric model, it should be a piece of cake to go from here to a fully accurate drawing.

As suggested, I drew the outline of the part as a polyline, then filleted the corners.

I could not find a way to input the thickness of the part only once. And, as this part is manufactured as a stamping, I did not want to place any measurements from a sharp corner to either the inside, or outside, edge of a fillet. Similarly, as this part is a stamping, the exact dimensions/centerpoint of the radii of the rounded corners are not critical to the function of the part.

I tried to keep dimensional constraints to a minimum, and utilized a lot of “equals” in this sketch, including the angles of the arcs.

Again, thanks to everyone for your help.

[kisolre]

You again forget that the fact that two points have the same coordinates does not mean that they are "COINCIDENT"ly constrained. The leftmost three centers of the inner and outer arcs are at he came spot but are not COICIDENTaly constrained. That is why you need several thickness dimensions. Box select the two centers (drag a square over them) and apply COINCIDENT constraint - the one in the first post of page 3 of the thread.

[chrisb]

Instead of repeating the measures you can use equality on the arcs.

[Gray Area]

You again forget that the fact that two points have the same coordinates does not mean that they are "COINCIDENT"ly constrained. The leftmost three centers of the inner and outer arcs are at he came spot but are not COICIDENTaly constrained. That is why you need several thickness dimensions. Box select the two centers (drag a square over them) and apply COINCIDENT constraint - the one in the first post of page 3 of the thread.

You are correct.

However, in my particular case, the thickness of the part is a variable I want to be able to change. I do not particularly care about the exact locations of the center points of the radii of the bends. In addition, the radii of the three (3) pairs of bends may, in fact, all be different.

I have made one change to my sketch. I deleted the 1.375 in. distance constraint and substituted a .125 in. thickness constraint to the middle vertical straight sections.

With this change I can go into the sketch, change the four (4) thickness constraints, and have the part change exactly as I want it to change.

Thanks again for your help and advice.

[Gray Area]

Instead of repeating the measures you can use equality on the arcs.

You are correct.

I did that from side to side. I did not, however, assume that all three (3) types of arcs are equal.

When it warms up a little more, I go out to the garage, dig around, find my part, and use a measuring tape and calipers to verify the part's exact dimensions. If all of the fillets are equal, I'll make that change to the final sketch.

Thanks again for your help and advice.

[openBrain]

Just for improvement, your design seems suboptimal.
There are a lot of tangency constraints missing where arcs join lines.
Also I guess I see tangency constraints on the arc centers, which seems quite unusual.
Last clue is that if your part in constant-thick, you shouldn't have to repeat the thickness and arc diameters.
If you could post your file, ones here may propose more optimal ways.

EDIT : maybe the 2 first statements in italic are a display effect...

[Gray Area]

Just for improvement, your design seems suboptimal.
There are a lot of tangency constraints missing where arcs join lines.
Also I guess I see tangency constraints on the arc centers, which seems quite unusual.
Last clue is that if your part in constant-thick, you shouldn't have to repeat the thickness and arc diameters.
If you could post your file, ones here may propose more optimal ways.

EDIT : maybe the 2 first statements in italic are a display effect...

Those tangent constraints are all there.

Here's a post of the file as it is to date.

Thanks.

[openBrain]

Attached my proposal.
As said, most "problems" come from missing coincidences between the arc centers.
From my PoV, it is cleaner now because :
* There are remaining only 1 overall dimension and 1 offset dimension on each axis
* Thickness is defined only once
* Radius of "plies" is defined only once
* There are lot less "equality" constraints (that was strange that you had to set equalities between inner & outer lines / inner & outer arcs)

It could be discussed whether equality constraint #53 should be replaced by a point-on-line constraint between the top right point and the Y axis.

HTH

[Gray Area]

Attached my proposal.
As said, most "problems" come from missing coincidences between the arc centers.
From my PoV, it is cleaner now because :
* There are remaining only 1 overall dimension and 1 offset dimension on each axis
* Thickness is defined only once
* Radius of "plies" is defined only once
* There are lot less "equality" constraints (that was strange that you had to set equalities between inner & outer lines / inner & outer arcs)

It could be discussed whether equality constraint #53 should be replaced by a point-on-line constraint between the top right point and the Y axis.

HTH

That is an elegant solution! I'm going to have to remember, and practice your method.

Thank you very much.

[openBrain]

That is an elegant solution!

Like that term. Artwork guys may show off, but technical elegancy is also enthusiasmic.

I'm going to have to remember, and practice your method.

Remembering is good, but practicing is definitely the way to go.