You don't need to be an 'investor' to invest in Singletrack: 6 days left: 95% of target - Find out more
https://photos.app.goo.gl/6p5uw6fEhgmAPQTo9
Anyone have any info? The photo was doing the rounds on FB lately.
I can never unsee that, hideous, should come with a warning 😳
It's an interesting idea for sure. But why that design!? Frankly it's offensive 😆
It looks like it has been designed via an AI like structural software. The software uses FEA/AI to come up with the optimal design for your requirements. Not pretty but potentially effective.
Yeah sorry I should have said, it's as much the design as the manufacture that's interesting to me. Putting material only where it is needed.
I've been doing topology and multi-disciplinary design and optimisation for about 12 years and for additive manufacturing for about 10 years.
The approach is all well and good, but you have to ensure that you capture every load, every combination of those loads and each potential perturbation of those loads and their combinations. Also, there is some anisotropy in metallic AM, not much but it's there and is far more common on poorly controlled machines and materials. You either need to capture or control all of that ^^ or place a sizable reserve on either the loads or material allowables lest it fail in use.
As for the specific design above - it's absolutely bloody appalling. Almost no post processing has gone into that design, they've thrown it at the software, smoothed (globally) what the software output was (which is highly influenced by the grid and the parameterisation) and then slapped it into a printer.
Really it should go:
1. Load and material understanding
2. Design space allocation
3. Coarse topology optimisation (TO)
4. Fine topology optimisation
5. Shape optimisation of the TO output
6. Proper design extraction and design for manufacturing (DFM)
7. Orientation selection (you might iterate between 6 and 7 here)
8. Support generation
9. Build simulation
10. Print
It looks to me like only 2, 3 and 10 were really done.
Looks like something you'd shoot someone with in a David Cronenberg film.
Also looks kind of like they put the crank arm end 170mm off the bed then turned tree supports on in cura.
Interesting Daffy, cheers.
Any ideas whether this could be better Thant he essentially tubular ish designs we have presently? I guess you'd need to see all the strain data...
Thomas Sandlerer on YouTube does a video about this kind of topology optimisation to create some 3D printed shelf brackets. It gives quite a good overview of the process.
100% what Daffy said. And I'd say step 1 is where the majority of the problems begin.
I've a little experience at the fatigue testing side of this. It rarely works on real components with more complicated load paths than shelf brackets. And even then, they'll have stress hot spots if the wall isn't perfectly flat or one rawlplug pulls a bit loose.
For cranks - no thanks (the optimisation side - the manufacturing process might be ok).
Now those are sexy Edd
Thanks @Northwind The cranks are the product of far too many hours and pounds. Super pleased with how they've turned out though.
The form of products is as much dictated by the options to manufacture them as it is by the design. Additivie manufacturing offers up an opportunity to free up designers significantly which is why if you design something with additive manufacturing in mind it will look very different to conventionally designed components that are often significantly restrained by the available and cost effective manufacturing methods.
the company I work for are experimenting with additive machining and are re-designing some existing components for testing and they look significantly different to the conventional components...much more akin to the crank at the top of this page...highly optimised designs al out organic in their appearance. The benefit of additive machining is not in the actual manufacturing process its in the design optimisation that is possible.
Biggest problem for us is consistency in material and mechanical properties from component to component is not there yet to sufficient repeatability and consistency which just drives you to over-engineer the component which significantly degrades the key benefit of additive machining in the first place. But that'll come.
@edd I really like your cranks. Will you be producing more of these in future?
Interesting concept and interesting discussion. I think someone has failed at step 1 - the asymmetry suggests that they are modelling downward pedal pressure, however when your pedals are level and you are standing on them then the load in the reverse direction on the crank is potentially much greater. I wouldn't want to land a jump on those.
@leegee I originally planned to run a Kickstarter campaign for a batch of ten sets to see whether there is a market for them. Things have changed though as I am about to start an exciting new job near Berlin. I don't think I'm going to have any time, so I guess I'm not building any more in the short term.
@edd, Guessing you had access to a machine through your work place? Not the sort of thing one has in one's garage.
Near Berlin? Berlin is a fab city
Molgrips - I did think about that, but isn't the structure based on parts of the crank.structure being either in tension or compression...and at 180° those roles are simply reversed?...and given incompressible nature of metal, will be ok?
Daffy's got a lot more experience in this precise sector than I have (humble mech. eng.) but this looks a lot like the sort of stuff that you'll get shown from Autodesk or Ansys, claiming that the software can iterate to a solution, if you define your inputs- (materials, boundary conditions, loading, available volume).
Looking at that spaghetti nightmare the designer has missed some really obvious loading information and has had no one with an engineering background check the solution.
Due to the fact that there is nearly no enclosed area in the cross-section, there is no torsional strength to the crank. Why do you think that lightweight cranks all have a hollow section through the crank arm? It's not because they had spare tubes that needed using. For a similar second moment of area a tube is stiff compared with an open section (like an I-beam).
VW did something like this for their new camper they processed the wheels and the wing mirrors. Got lots of talk at the trade shows but then their vans closer to production had far more normal components.
Why do you think that lightweight cranks all have a hollow section through the crank arm? It’s not because they had spare tubes that needed using. For a similar second moment of area a tube is stiff compared with an open section (like an I-beam).
As described a hollow structure (tube) is the best design for high performance cranks. Additive manufacturing allows me to a) make the cranks hollow and b) vary the wall thickness, as required, throughout the structure. The problem, from a marketability point of view, with my cranks is that you cannot see any of this...
The problem, from a marketability point of view, with my cranks is that you cannot see any of this
I was thinking that. You cranks look nice but they don't stand out from off the shelf items. All the cool engineering is on the inside. That'll be a hard sell as I expect the price is pretty high. You almost want to add some extra, unique external features but from and engineering point of view that would probably make them worse.
@nickjb Yeah, it's a difficult balance. As per my post higher up, my new job means that any effort to sell them is on hold anyway. Ultimately when I started the project I wasn't planning to sell any, I was just trying to make the best cranks possible.
148g for a crank is pretty cool (I may be biased!):