the easiest way to design this?

the easiest way to design this?

Anonymous
Not applicable
2,442 Views
27 Replies
Message 1 of 28

the easiest way to design this?

Anonymous
Not applicable

Hi guys,

 

I just designed this to repair my Hamilton Beach drink mixer.

 

While I loved the job, I think I didn't necessarily do it the best way so I was wondering on how you experts would go about it?

 

Anny suggestions for alternative ways to do it are welcome as we will always learn something new. i don't want to influence you so I will tell you how I did it afterwards.

 

curious to see your replies.

 

Best

 

Oliver

0 Likes
Accepted solutions (2)
2,443 Views
27 Replies
Replies (27)
Message 2 of 28

jodom4
Community Manager
Community Manager

Hey Oliver,

Do you have a photo of the thing you're trying to repair?


Jonathan Odom
Community Manager + Content Creator
Oregon, USA

Become an Autodesk Fusion Insider



0 Likes
Message 3 of 28

davebYYPCU
Consultant
Consultant

Ignore the hub / shaft section, 

I would sweep a surface, thicken and fillet.

 

In Tspline would distort a flat cylinder.

0 Likes
Message 4 of 28

chrisplyler
Mentor
Mentor
Accepted solution

My ideas...

 

 

 

Message 5 of 28

Anonymous
Not applicable

Excellent Chris,

 

Very smart solution.

 

Mine was much more complicated.

 

I created a disk and then on 90 degrees a cylinder extruded with an angle to which I applied a circular pattern and then used the latter to cut out the waves from the disk.

 

Then I smoothed everything with splines.

 

Quite complicated...

 

Thumbs up. I learned something.

 

Best

 

Oliver

0 Likes
Message 6 of 28

Anonymous
Not applicable

thans for replying. but I don't fully understand what you mean...

 

Best wishes

0 Likes
Message 7 of 28

davebYYPCU
Consultant
Consultant

I was too brief, but exactly the same construction as Chris did, with the exception that I make the centreline of his  "Slot", in Patch then thicken and fillet, 

his solid Slot Sweep is more efficient.

0 Likes
Message 8 of 28

chrisplyler
Mentor
Mentor

Thank-you for the kind assessment Oliver. It was a fun challenge.

 

I always try to think carefully about the required geometry and whether there is some way to create it without leaving the Model workspace, just using sketches and the basic features. I much prefer the mathematically precise nature of this as opposed to pushing/pulling nodes by hand/eye. In my opinion it's worth putting in a few minutes of thought towards this goal before giving up and resorting to Tsplines and stuff.

 

After my video I put in a bit more thought, and revised as follows:

 

1. Edited the sweep. Swapped the path and the guide rail. Set the path as the wavy spline and set the guide rail as the flat inner circle. Then set the Extent setting to "Perpendicular to Path" and set the Profile Scaling setting to Stretch. This combination means that the slot profile follows the wavy condition properly, resulting in a wave that can have a mic put on it at any position and measure a consistent thickness, because the slot profile pivots as it is swept along the wave, and because as it needs to get longer/shorter to reach the circle perimeter it is stretching only in length instead of scaling up the thickness also as it was doing before.

 

2. Added fillets.

 

mixerpart.jpg

Message 9 of 28

chrisplyler
Mentor
Mentor

 

So now I can do this....

 

 

 

Message 10 of 28

TrippyLighting
Consultant
Consultant

This can also be done easily with T-Splines:

 

 


EESignature

Message 11 of 28

lichtzeichenanlage
Advisor
Advisor

@TrippyLighting: Thx for the posting. As stated so often I haven't done anything in the sculpting environment. That doesn't mean I don't like to learn things 😉

 

What's the benefit of doing it this way instead of in the model environment. For me both ways starts somehow similar but your way looks more intuitive but less precise and the timeline is missing. Is the surface better? 

 

Edit: Typo removed

0 Likes
Message 12 of 28

davebYYPCU
Consultant
Consultant

Timeline is not missing.

 

Sculpt area is the same as Direct Modelling, and sketching, make the articles, without history,

when done Finish Button, puts one icon in the timeline.

 

Any of those three environment allow editing afterwards.

 

 

Message 13 of 28

TrippyLighting
Consultant
Consultant

@lichtzeichenanlage wrote:

@TrippyLighting: Thx for the posting. As stated so often I haven't done anything in the sculpting environment. That doesn't mean I don't like to learn things 😉

 

What's the benefit of doing it this way instead of in the model environment. For me both ways starts somehow similar but your way looks more intuitive but less precise and the timeline is missing. Is the surface better? 

 

Edit: Typo removed


 

That is a very good question!

the title was what is "the best" way to model this geometry. The matter of fact is that "best" often only exist in some specific circumstances.

 

The method used by @chrisplyler is definitely a good way to do this. But it does require some setup of sketches, splines, etc. And it is parametric, which is something I like.

 

However, I wanted to simply present a different way to create very similar geometry. The notion that T-Splines are not "precise" is very prevalent amongst traditional CAD people. Most CAD people never do anything but solid modeling and there are clear limits to that! One of these limits is surface quality in terms of curvature continuity. That is unlikely a requirement for this object, however, T-Splines are naturally G2 curvature continuous except at poles (vertices that do not join 4 edges, so either 3 or 4+n).

in this case there are no poles to the whole thing is G2 curvature continuous.

T-Splines are NURBS compatible and can be every bit as precise as a NURBS surfaces. There is a reason many AD competitors have developed tools similar and in some cases faster and more powerful that T-Splines so a quad-mesh model can be transformed into a CAD accurate solid model.

 

Another reason for me to work with T-Spliones is that you can very quickly explore forms. This shape can be very quickly edited  into another form, which is unattainable with a sketch based solid, or surface model. If Fusion 360's T-Splone modeling tools would not be so clunky and functionally deficient I would probably use Blender less often, but the combination of Blender in conjunction with T-SPlines is very powerful.

 

Also, as @davebYYPCU mentioned, there is no timeline. Particularly when developing complex shapes (and not only there) the timeline can become a real burden.  


EESignature

0 Likes
Message 14 of 28

chrisplyler
Mentor
Mentor

Trippy gives a VERY good answer. "Best" depends on the criteria.

 

Let's consider shape. You, Trippy and I have all produced a model. All three will probably agitate a liquid when you spin them. Given the right size and the right bore hole to fit the shaft in your mixer, all three of them would probably work fine. Since you are repairing your mixer, it might be a decent idea to look at how the previous part failed, and if a slight design change might improve longevity of the replacement part you design. For example, if the wavy bit stripped off of the center bit, then you might want to add the fillets at the join as I have done above, and perhaps you even want to thin the blade profile as the diameter increases so that there is less mass farther out.

 

But suppose it was a part for some kind of pump on an aircraft. The shape of the wavy blade might be of critical importance, for precisely calculated flow numbers at a precise rpm under a given load. And not only flow rate, but the design and material must stand up to that rpm and load resistance for so many millions of revolution cycles. Okay so an engineer specializing in fluid dynamics and another specializing in mechanical engineering have determined the exact shape of the wavy bit that is required, and now YOU have to model it.

 

Which Fusion tools can deliver the exact shape you need with the least amount of effort, or with the ability for parametric changes, or blah blah blah, or blah blah blah, etc.? Only YOU can decide.

 

Suppose you DID decide to thin the blade as diameter increases. Can you do that easily with your own developed model? I can just redraw that slot I used as a sweep, making it narrower at the outside. And I can define the taper exactly with an angle if I want. Trippy can just move a couple rings of Tsplines closer together. I don't know enough about the sculpting environment to know whether he can define a precise angle like I can...? Trippy can you elaborate on this?

 

I'm self-taught in Fusion, and like you I haven't really explored the sculpting environment/workflow much at all. In order to decide which method is best, we're going to have to put some time into learning all the possible methods so that for any given project we can assess the pros/cons of each method.

Message 15 of 28

lichtzeichenanlage
Advisor
Advisor

@TrippyLighting: Thanks for the very good and detailed answer. I'll guess I have to spend some time in the sculpting environment but my bett is, that I never gain good skills in this.

 

@chrisplyler@davebYYPCU@TrippyLighting: I hope nobody felt my question kind of rough against @TrippyLighting. Perhaps it's because I'm from Northern Germany. We sometimes work like that.

0 Likes
Message 16 of 28

Anonymous
Not applicable

Hi guys,

 

great discussion, thank you to all. I will have to look at the alternatives.

 

As the question to what the issue was with the part, let me explain and add some pictures to illustrate.

 

1) the original (white) part is not round but star shaped. if it hits protuberances present in the glass, it can break which is what happened to my part as you can see in the picture.

 

2) the material strength is the same everywhere, so it lacks strength in the middle where it is connected to the shaft

 

3) I have another mixer which yields much better results. it has two disks instead of one and the disks are round. I guess this is better to create small bubbles in the shakes as the sharper star form with hard edges may damage the bubbles (I supposed that).

 

so the idea was the following:

 

1) produce a round part that would not easily be damaged when hitting the protuberance

 

2) put two or more instead of one

 

3) reinforce the middle

 

my design achieved all this and the shakes I made with it are very creamy, but is much less elegant than some of your designs.

 

I have come up with another way of doing it but have not yet finished checking if it is OK when transferred to a 3D print file.

 

best wishes

 

 

 

 

0 Likes
Message 17 of 28

Anonymous
Not applicable

I think for the purpose, this is the easiest way as Chris suggested:

 

"1. Edited the sweep. Swapped the path and the guide rail. Set the path as the wavy spline and set the guide rail as the flat inner circle. Then set the Extent setting to "Perpendicular to Path" and set the Profile Scaling setting to Stretch. This combination means that the slot profile follows the wavy condition properly, resulting in a wave that can have a mic put on it at any position and measure a consistent thickness, because the slot profile pivots as it is swept along the wave, and because as it needs to get longer/shorter to reach the circle perimeter it is stretching only in length instead of scaling up the thickness also as it was doing before."

 

inverting the path and the rail enabled me to reproduce where at first I was always getting the dreaded intersect error.

 

I attached a file with a variable section slot I used.

Message 18 of 28

TrippyLighting
Consultant
Consultant

@Anonymous What you are doing is really product design. Sometimes companies are very thankful for feedback. If you send them your findings and changed design, they might send some free products in return.

 

@lichtzeichenanlage no I did not perceive your question as rude whatsoever. I am German myself (Grew up on the outer fringes of the Ruhrgebniet, close to the Dutch border ) and can be somewhat rough around the edges. Call it German ruggedness 😉

 

 


EESignature

Message 19 of 28

chrisplyler
Mentor
Mentor

Very nicely done Oliver.

 

Like I said I've barely ever played around with the sculpting environment, but having been inspired by Trippy, I created this in about five minutes....

 

mixerpartsculpt.jpg

 

I'm very happy with it. But I can't tell you what the radius is around the top of the cylinder, or whether it actually is a constant-radius arc fillet at all. I can't actually tell you many details about the shape at all, other than its outer diameter and height, because it was all dragged into place on my screen just by eyeballing it. As an artist, I would think I had done the job. As an engineer, I would be at a complete loss. My mind is most comfortable operating in the concrete, definitive world of sketching exactly what I want.

 

@TrippyLightingyou said that sculpting didn't have to be imprecise. Can you offer a bit of insight into how one might more definitively control the sculpture? Creating that mixing head was very easy, and has opened my eyes to exploring this whole sculpting world. But suppose you were copying an existing part exactly, and you had your calipers out taking measurements of thicknesses and radii and stuff, how do you establish same in your sculpted shape?

0 Likes
Message 20 of 28

chrisplyler
Mentor
Mentor

I think making the root three or four times thicker than the edge and filleting it as well will probably prevent it ever cracking off down at the root!

 

mixerpartsect.jpg

 

GOSH DURNIT all you morons are forcing me to go buy a drink mixer and a 3D printer. DANG YOU ALL TO HECK!

 

0 Likes