CAD Forum

I am trying to make this rib follow the edge of both cylinders, but no matter the approach (using features) I never get a clean look.
How would you do it?
I did it using surfaces, but I guess there is a easier way:

I could not open your file due to version diff
But based on my understanding you would like to have a rib that go around the cylinder surface

If it is a simple part, I would try to loft
1. Split line 2. Loft Feature tree: Side note: I am using a top plane sketch to create a split line instead of a right plane sketch so that it will not create unwanted tangent edge at the opposite side of cylinder

The easiest way to do it is to draw a short line at the ends of an angled line. The short lines are coincident with the tops of the cylinders and the angled line is the top of the rib.

I'm on my phone now, I'll add an image when I'm in front of the software.

Hi Lucas

There are multiple solutions, and some decisions to make when making design like this.

Is it required that the rib ends on the exact top faces of the cylinders?
Is it OK if we have a short straight wall from both cylinders, before making the rib?
Do we want to have it follow the exact curve on both cylinder top faces, making the rib top change curve direction?

I would allow the rib to start a bit under the cylinder top, and then, if needed, add a fillet to blend it smoothly.

Not sure how to attach a file here....

Zhen-Wei Tee wrote: ↑Fri Jun 25, 2021 6:58 am I could not open your file due to version diff
But based on my understanding you would like to have a rib that go around the cylinder surface

If it is a simple part, I would try to loft
1. Split line
image.png
image.png

2. Loft
image.png
image.png
image.png


Side note: I am using a top plane sketch to create a split line instead of a right plane sketch so that it will not create unwanted tangent edge at the opposite side of cylinder

A practical method,
Hi @Zhen-Wei Tee
I recommend,

if there are feature tree images, those who do not know will understand the method better.

Tim R. Halvorsen wrote: ↑Fri Jun 25, 2021 7:31 am Hi Lucas

There are multiple solutions, and some decisions to make when making design like this.

Is it required that the rib ends on the exact top faces of the cylinders?
Is it OK if we have a short straight wall from both cylinders, before making the rib?
Do we want to have it follow the exact curve on both cylinder top faces, making the rib top change curve direction?

I would allow the rib to start a bit under the cylinder top, and then, if needed, add a fillet to blend it smoothly.

Not sure how to attach a file here....

When you are in Full Editor & Preview mode,
Click the Attachments and upload the file

Ömür Tokman wrote: ↑Fri Jun 25, 2021 8:06 am A practical method,
Hi @Zhen-Wei Tee
I recommend,

if there are feature tree images, those who do not know will understand the method better.

Thanks for the suggestion, i had edit my post and include the feature tree image

@Zhen-Wei Tee Thanks, I was thinking if I was supposed to do it with Loft or Boundary, guess I need to practice both tools more cause it couldn't imagine how.

@matt The way you said looks like the way I did.

@Tim R. Halvorsen Well, I don't need those ribs there. Actually this is just an exercise, I found this blueprint in a Blender tutorial I was starting, and I tough to myself "This simple part is taking about 1h of tutorial? Wtf hahaha, I can do it in 2min with Solidworks, hold my beer." Then the irony kicked me hard.

This is the way I'd do it. The end points of the angled line have a Pierce relation with the edges of the cylinders, and it doesn't matter how long the short horizontal lines are as long as then end on the top faces of the cylinder.


It doesn't make the the intersection of the rib and the cylinder perfect, for that you'd need a loft, but then the top of the rib wouldn't be flat. Not that it needs to be, but avoiding unnecessary curves will probably keep a machinist off your back.
Without the short horizontal lines, the width of the rib doesn't have a face to stop it, so it would continue out into infinity. That's why it fails with just the angled line.

matt wrote: ↑Fri Jun 25, 2021 8:52 am This is the way I'd do it. The end points of the angled line have a Pierce relation with the edges of the cylinders, and it doesn't matter how long the short horizontal lines are as long as then end on the top faces of the cylinder.

It doesn't make the the intersection of the rib and the cylinder perfect, for that you'd need a loft, but then the top of the rib wouldn't be flat. Not that it needs to be, but avoiding unnecessary curves will probably keep a machinist off your back.

Thank you, that is how I did with the rib feature too

Thinking about manufacturability it is indeed the best approach

Lucas wrote: ↑Fri Jun 25, 2021 3:45 am I am trying to make this rib follow the edge of both cylinders, but no matter the approach (using features) I never get a clean look.
How would you do it?

image.png

I did it using surfaces, but I guess there is a easier way:

image.png

Maybe someone else brought this up but I didn't see it other than "Think about manufacturability".

How that rib is modeled will depend largely on how the part is going to be made. Is it a machined part, casting, Machined casting, weldment, plastic injected?

You're probably going to model it entirely differently depending on these factors.

MJuric wrote: ↑Fri Jun 25, 2021 12:42 pm Maybe someone else brought this up but I didn't see it other than "Think about manufacturability".

How that rib is modeled will depend largely on how the part is going to be made. Is it a machined part, casting, Machined casting, weldment, plastic injected?

You're probably going to model it entirely differently depending on these factors.

In my world, it is either machined into steel mold, or machined into a carbon electrode, which then erodes the steel mold.

In any case, it will probably be filleted all to hell and back, so the flat top may or may not be much of an issue in the end.

matt wrote: ↑Fri Jun 25, 2021 1:16 pm In my world, it is either machined into steel mold, or machined into a carbon electrode, which then erodes the steel mold.

In any case, it will probably be filleted all to hell and back, so the flat top may or may not be much of an issue in the end.

That's kind of why I mentioned it. We do castings, weldments and machined parts mostly. How I would model that part would be dependent on which of the three it was and then with the castings and weldments whether that part was involved in a machining process or not.

A plastic injected part, Sinker or other part is going to be modeled differently than any of these. Heck, it might even be designed differently depending on whether it was investment, sand or die cast.

MJuric wrote: ↑Fri Jun 25, 2021 3:58 pm A plastic injected part, Sinker or other part is going to be modeled differently than any of these. Heck, it might even be designed differently depending on whether it was investment, sand or die cast.

In injection mold, I've stopped second guessing mold makers. If we don't talk beforehand (and usually we don't because the mold maker speaks Chinese and works 12 time zones away) and I don't have any clue about what processes he's using, I won't pretend to force his hand. If it's important to him I would hope he'd come back to me. Chances are the top of a rib is in the +/- half the width of the feature sort of range. On a plastic part, it's probably an interior feature with nothing mating up to it, so even +/-0.02 isn't going to cause any heartburn, and a little arc on the top would certainly be in that range, especially if the corners are filleted.

When things like that are important, I'll often send an actual 2D drawing along with the 3D data (I know you machine design guys are probably mortified) just to point out the areas where "better than visual" tolerance is needed.

matt wrote: ↑Fri Jun 25, 2021 4:12 pm When things like that are important, I'll often send an actual 2D drawing along with the 3D data (I know you machine design guys are probably mortified) just to point out the areas where "better than visual" tolerance is needed.

Not at all. "Better than visual" is a way better method of tolerancing than putting +/-.005 when "Better than visual" will do.

I work with tolerances that run the gambit from +/-.0001 to "Anything smaller than this". WE don't even bother to tolerance our casting drawings, which I think is insane, but when you're dealing with a part that is essentially 4 ft cubed and you're making a mold like you would a sand castle and just filling it with molten metal...well you don't expect much :-)

Tolerance is tolerance as long as it's appropriate it is what it needs to be.

I'd do it like this to keep the round top face of the cylinders - but there are so many solutions available

@Lucas - It is not always easy to model simple stuff

As a machinist, sometimes programming what is an apparently easy part is much more complicated compared to what is considered an advanced part.


I'll add a new part for you, containing all the solutions I'd use on similar design.
There are 5 in total, and you can unsuppress/suppress 1, 2, 3, 4, and then 5.0-5.2.

Solution 1-4 are single feature solutions and the 5th is a 3 feature solution for the rib.

Roll-back and use the model and create relations - use the 'show temporary axes' to create collinear relations, to have design update with design changes.

Tim

MJuric wrote: ↑Fri Jun 25, 2021 12:42 pm Maybe someone else brought this up but I didn't see it other than "Think about manufacturability".

You're probably going to model it entirely differently depending on these factors.

I am working with additive manufacturing, so in my case some details like this are more important due to aesthetics ^^

Tim R. Halvorsen wrote: ↑Sat Jun 26, 2021 2:59 am @Lucas - It is not always easy to model simple stuff

Thank you Tim, the rib a little lower actually looks much better