The huge disadvantage of generative design is. If your input is somehow incorrect or you have forces you didn´t recon with your part will fail. You have basically zero error margin. Additionally the parts can only ever be produced with additive manufacturing. So if you would even want to make a cheaper part out of common materials like sheet metal or if you want to create a series of parts not a one off this would mean you need to completely redesign everything again. It´s a neat tool for rocket engineers or aviation parts because it saves max. weight but outside of this it´s just cool looking. Often small features suck as well. Like in your case at the biggest round mounting point, there is a small branch. You would never make a part like that, because this small branch adds complexity to manufacturing which does not need be, you can just change the design in this area to get rid of that gap it creates. It prints faster this way and you can create a mold or similar much easier. So yeah nice toy to play with and looks futuristic but one need to know it´s downsides as well. The price for the SLM part is pretty good though. It usually cost bout 150€ for me for a similar sized object, so I am pretty surprised how cheap it has become.

@bubsbuilds

12 күн бұрын

Definitely agree that it's unlikely we'll ever see additive being the tool of choice for any high production parts, and that generative design tools to date can't seem to produce parts well-suited to other options. But disagree a bit on that opening 'disadvantage' being something unique to generative design...although it can do a hell of a job of highlighting those oversights! I think that's the case for anything being 'engineered' for purpose, especially if that thing is going to be produced in any volume. That 'margin' is coming out of somebody's bank account in material costs, no matter what process is being used to make the parts. It's that trade-off of upfront engineering costs vs. long term material costs...that damn Coke can has had some bucks spent on it! And that margin can be had with generative as well, just crank up that 'Safety Factor' number and let it go to town. But couldn't agree more that generative design seems like a bad choice for prototyping something out of sheet metal! And likewise for anything where it's not worth the upfront effort, since generative is indeed much less forgiving on designing without good part requirements. If nothing else, I think it could be a great training aid for young engineers, then maybe I won't end up throwing out so many poorly designed injection molded parts! Oh, and I would make a part like that, and I did :) Very much appreciate the thoughtful, detail-rich addition. May have a couple of "agree to disagrees", but lots of really good points, thank you!

@thomas8719

11 күн бұрын

Take a look at "Synera" if you want to see where the future might be heading. This tool addresses many of the problems you describe.

@MRM.98

11 күн бұрын

You’re spot on. DFM becomes problematic. Cool to look at and design but not very practical.

@sierraecho884

10 күн бұрын

@@bubsbuilds I think we have more of a misunderstanding situation than disagree situation. Let me explain. With the error margin I din´t mean that you can´t simply crank up the thickness a little more I mean if you only design for a force in X and Y direction because you think that´s all that need be but due to vibration, temp. humidity, unforeseen events etc. the part need more load bearing in Z as well you have no margin. But you are right this can happen with a traditional design as well, so there shouldn´t be much difference except ease of manufacturing for high volumes and price without generative design. Injection molded parts are usually also 3D printed first, well at least that´s what I do and all my co-workers before we commit to that $100k mold. But yeah, as long as you are aware of it´s limitations it should be fine. On the other hand it´s pretty cool looking (could be great for marketing), it saves a lot of material and you can manufacture it easily via additive methods. Mercedes has done something similar lately on one of our latest projects, they molded a whole section on their car with one piece and skipped the 5-8 sheet metal parts which needed to be joined. They then made sure they can cast the part out of metal instead of producing it by additive manufacturing and voila pretty cool looking material saving mass production ready part. So it has it´s purpose for sure. In the end one can argue it´s even a better process. Usually I create a design first from past experience. Then comes the FEM-analysis and then countless interations of design changes. Here it´s the other way around. Generative design starts with the FEM-analysis and let´s me choose the parts, it already creates a couple of iterations beforehand. I am not sure if it predicts the non isotropic behavior of FMD printed parts though, but that depends on the software you use and the margin you choose so it´s solvable. So all in all I have to agree you made a good point, the issues I pointed out can happen with traditional designs just as well.

@sierraecho884

10 күн бұрын

@@thomas8719 Thank you I will check it out.

As a software engineer spent my whole life in coding, I always admire people who can build the real things. I am also thinking about buying a 3D printer and learn how to build my own stuff. Bump into this video and really amazed by how versatille you're in terms of building all the stuff yourself 🤯 Just a bit curious what you do for living and how did you learn these skills?

@bubsbuilds

10 күн бұрын

If it makes you feel any better, I've spent much of my life wishing I could write some half-decent code :) But thank you, that is really quite the compliment, and I love to hear that you're interested in expanding to hardware!! I am a mechanical engineer, by both education and current job title. As to the 'how', hmm, I can do my best guess, but please add a 'grain of salt' since it's purely a guess from self-reflection here. I think part of it comes from a combination of being naturally curious and having that same sort of desire that you do about building hardware. But the biggest part, in my opinion, has been the incredible good fortune of crossing paths with a lot of people MUCH smarter than myself that were gracious with their time and attention. There are few things I enjoy more than being the dumbest person in a room, and I've found that if I embrace my ignorance in them, many of those folks are more than happy to indulge my curiosities. To be honest, that's a huge part of why I decided to try sharing the projects I dive into in my spare time. I by no means have the depth of knowledge or skill of those I've pestered with questions over the years, but my hope with this channel is very much to be able to pass along at least SOME of the things they've been kind enough to share with me along the way...and hopefully also convey my passion and gratitude for it as I do. Sorry, very long-winded and non-specific response :) But in short, get the printer, try the thing, ask the questions...just like you already are! I'm happy to try my best to answer questions along the way (if I can), but I may ask you to review code snippets in exchange!

Great video! I have been looking into generative design and curious how to get started. What is the software you use, is it a Fusion360 plugin?

@bubsbuilds

12 күн бұрын

Thanks! And yep, I used the Fusion360 "Simulation" Extension (they bundled the GD stuff inside of the sim package at some point.) Unfortunately it's VERY pricey to buy a license, but they have a two week trial for it. If you do decide to give it a go, I'd be curious to hear how it goes for ya. And if you run into any questions or problems, I can't promise I'd be of any help, but feel free to send em my way!

👍

Excellent video. Nice to see what came of your design work in the last video. Can we expect a follow up with the load tester to see if you get closer to predicted loads? It looks like the time lapse images were very stable. You should really try a titanium based design for a Bamboo labs running at full speed ;)

@bubsbuilds

10 күн бұрын

Why thank ya, good sir! Yep! I bought myself some steel stock to make some more rigid fixturing. My plan is to upload some follow up tests as shorts...although it'll depend on when I get the chunk of time to turn that steel stock into something useful. Do I get to attach it to the bed on the Bambu? If so, I think I like my odds :) .... but guessing that's not quite the challenge you had in mind. Although hell, with their machines, I'm not sure my GoPro has the frame rate to keep up!

Should try another design with the camera leaning forward towards the print, while making the bracket longer so the camera's weight is closer to the mounting point. Less deflection with bed movement

@bubsbuilds

7 күн бұрын

@yousefosman8094 the current set of mounts are pretty stable. When intentionally cranking up the loading, like when I put my full-sized camera on it, there seems to be more deflection coming from the Prusa Y-axis carriage than from the mounts...which I suppose makes sense, given that the carriage is a narrower aluminum cross-section than the mounts in places. So I'm not sure that would buy much in performance, and it may make the fan clearance issues a bit worse. But if you have a specific offset and angle combo you'd like to see, I'd be happy to run it and upload the model for ya! I'd just need 1) the offset from the bed edge in Y, 2) the height offset from the bed in Z, and 3) the pitch angle, thetaX, for the camera mount. Cheers!

Cool video, mate. This channel is becoming a goldmine, and I'm here for it!

@bubsbuilds

12 күн бұрын

Thanks! Very much appreciate it, and I'll see what else I can dig up :)

Excellent example of over enginnering! Love it. 😍

@bubsbuilds

12 күн бұрын

You get me :)

Music is unnecessarily loud and thus the video is harder to understand - at least to (my) not native english ears

@bubsbuilds

10 күн бұрын

Thank you, that sort of feedback is VERY helpful! It's hard for me to gauge, since by the time I'm adding music I've heard the words a good many times already in editing :) Unfortunately there's no way for me to fix it in this one (YT doesn't allow for 'revised' uploads, as far as I know). But it definitely with help me in setting levels going forward. Really appreciate it!