New crank design

[AMuzi]

https://bikerumor.com/2021/05/21/sram-prototypes-ai-optimized-3d-printed-cranks-using-autodesk-generative-design/
--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

[Tom Kunich]

On Thursday, May 27, 2021 at 1:07:59 PM UTC-7, AMuzi wrote:
> https://bikerumor.com/2021/05/21/sram-prototypes-ai-optimized-3d-printed-cranks-using-autodesk-generative-design/

The problem as I see it is that 3D printing is not a process without errors. It can very easily have voids and weak spots.

If you had the idea of using the AI for designing the lightest possible part supposedly with the highest strength per weight ratios, what does it matter? Most of the bikes in the Tour will be too light for the rules and have to carry compensating weights.

If the UCI, like expected, eliminates the weight limits it would probably be the end of high end cycling because it would be like lighting a fuse on a stick of dynamite. Anyone that is close to it will be injured by the flying debris.

[AMuzi]

Yes and no.

Many 3D print pieces do have some severe limits where
strength, torsional resistance or pressure matter. But
sintered pieces from 3D metal powder are failry well known
and of a different class. I don't know but it could well be
adequate and reliable.

[Tom Kunich]

https://3dprinterly.com/are-3d-printed-parts-strong/

[Steve Weeks]

On Thursday, May 27, 2021 at 5:54:44 PM UTC-5, cycl...@gmail.com wrote:
> https://3dprinterly.com/are-3d-printed-parts-strong/
Thanks for the interesting article. I had my broken ankle screwed back together with screws made of polylactic acid ("PLA") three years ago. I expect they were 3-D printed. Not only were they strong enough for the job, but they were resorbed after the joint ("tibio-fibular syndesmosis", if you *must* know!) healed. So no second surgery to remove them!

[Frank Krygowski]

On 5/27/2021 4:07 PM, AMuzi wrote:
> https://bikerumor.com/2021/05/21/sram-prototypes-ai-optimized-3d-printed-cranks-using-autodesk-generative-design/

Pretty wild looking! A few things I wonder about:

Exactly what loads did they use for inputs to the software? Did the list
include forces from unplanned events, like crashing and loading the
crank sideways? Or how about something really random, like whacking your
crank against a rock, which might damage one of the strut members? I
could envision such a load bending a small strut and making the crank
look fine, but be far weaker.

I wonder what results they got for mass and stiffness, and how it
compared with more conventional cranks? Are the gains worth the cost and
complexity?

Will they put a skin over the crank to keep it from filling with mud?

And: It always seems to me that even normal loads on bike parts are
difficult to predict. They can even depend on design - for example, when
a mountain biker lands a big jump, the peak force will depend not only
on the jump and the mass and skill of the rider, but on the stiffness of
the suspension (if any) and the stiffness of the crank.

So how are they evaluating the prototypes, to validate the computer
data? Seems they'd need some experimental stress analysis - strain
gages, brittle coatings, etc. It's not simple.

But it's an interesting project.

--
- Frank Krygowski

[AMuzi]

I, like you, have seen a lot of things and seen a lot of
things fail. I agree that modern 'modeling' seldom accounts
for real world outliers with relatively high probabilities.
Which is to say 'unexpected' describes the designer more
than the event.

[Axel Reichert]

Frank Krygowski <frkr...@sbcglobal.net> writes:

> Pretty wild looking!

Many topology optimization results look much wilder.

> Exactly what loads did they use for inputs to the software? Did the
> list include forces from unplanned events, like crashing and loading
> the crank sideways?

I cannot comment on SRAM, but I do know something about Trek:

https://www.3ds.com/fileadmin/PRODUCTS/SIMULIA/PDF/case-study/SIMULIA-trek.pdf

Adaptive manufacturing becomes really high-end once you start to
simulate the process itself (which you should, but I am biased, since I
am working for this company):

https://blogs.3ds.com/simulia/print-to-perform-overview-of-additive-manufacturing-on-the-3dexperience-platform/

Best regards

Axel

[Sir Ridesalot]

That looks to be an accident waiting for a place to happen. To me it looks like a very weak design. I wonder too what'll happen when the bicycle goes over obstacles or if there's a twisting force on the crank.

Cheers

[Rolf Mantel]

I've read somewhere that the aricraft industry is using 3-D printed
Titanium joints which weigh about half as much but are double as strong
as the previous generation of joints.

Rolf

[AMuzi]

On 5/28/2021 3:12 AM, Sir Ridesalot wrote:
> On Thursday, May 27, 2021 at 4:07:59 p.m. UTC-4, AMuzi wrote:
>> https://bikerumor.com/2021/05/21/sram-prototypes-ai-optimized-3d-printed-cranks-using-autodesk-generative-design/

>

> That looks to be an accident waiting for a place to happen. To me it looks like a very weak design. I wonder too what'll happen when the bicycle goes over obstacles or if there's a twisting force on the crank.
>
> Cheers

> "I wonder too what'll happen..."

http://www.yellowjersey.org/photosfromthepast/CRANKMD9.JPG

[AMuzi]

+1

3D additive machining is much more than little plastic gizmos,

[Ralph Barone]

On the other hand, if you chose to, you could keep the weight the same and
design a stronger crank using these methods.

[Tom Kunich]

I think that the AI and 3D printing of stronger cranks is a perfectly valid way to model things so that you could use the vastly complex methods of casting such pieces.

When we were going through that business about the steel rotation limiter, I pulled an aluminum casting off of a cheaper Campy derailleur and tried it and it was too thick and could not work. But when I ordered another from Banford Bike, I ended up getting another aluminum casting. I wrote to them three times and never got an answer. I KNOW that they still make this steel piece because all of the 11 speed derailleurs use it. But eventually I tried the aluminum casting and it was indeed different and it appears to be able to work. So I installed it and hope that the Centaur derailleur doesn't fail again.

I could always install a Record derailleur if I get too nervous.

[Frank Krygowski]

Right. And some of the biggest development efforts occur about seven
miles from me. https://www.americamakes.us/


--
- Frank Krygowski

[Fit Printers]

I appreciate the informative information. Three years ago, I had polylactic acid ("PLA") screws used to repair a broken ankle. They were probably printed using 3-D technology. In addition to being capable of doing the job, they were also resorbed once the joint (called the "tibio-fibular syndesmosis" if you *must* know!) healed. Therefore, there is no need for a second surgery to remove them!

https://fitprinters.com/best-4x6-photo-printers/

[Lucyharper]

> https://rediyart.com/best-printer-for-stickers/

[SaaS Alternatives]

Thank you for the useful information. I had polylactic acid ("PLA") screws put in three years ago to fix a broken ankle. They were most likely created utilizing 3-D printing. They were not only capable of doing the task, but they were also resorbed after the joint, known as the "tibio-fibular syndesmosis," healed. Therefore, a second surgery to remove them is not required!

https://saasalternatives.net/puppeteer-alternative/

[Tom Kunich]

These are all 2D printers which have nothing in common with 3D printers.

[Tom Kunich]

Again, these are 2D printers.

[Tom Kunich]

I prefer the term "absorbed". That material wasn't there to begin with so "resorbed" is a false term though perhaps MD's use it because that was the way it was marketed to them.